«Scenarios» Section#
Scenarios are necessary for generating events and incidents. A handler must be assigned to the scenario. Scenarios and handlers are configured in the «Scenarios» section.
The general view of the «Scenarios» section (Figure 42).
The main interface element of this section is the table of scenarios. The table includes the following columns:
- Name – the name of the scenario;
- Type – scenario type (event or incident);
- Number – the number of recorded registrations (events or incidents);
- Lists – a list with which the attributes of the vehicles and LP are compared;
- Handlers – the name of the handler used for the scenario;
- Creation date – the time and date the scenario was created;
- Status – the activity status of the scenario.
Search and filtering are possible by the columns «Name», «Creation Date», and «Status».
Scenarios can only be created, filled, modified, and deleted by the Administrator.
Creating and Editing a Scenario#
To create a new scenario, click the «Add» button.
General view of the scenario creation page (Figure 43).
To edit a scenario, click on 
in the scenario row, and the scenario editing page will open (Figure 44).
The description of filling out the creation and configuration page of scenarios is provided in Table 13.
Table 13. Scenario options
| Name | Description | Values |
|---|---|---|
| General | ||
| Name | Field for entering the scenario name | Can consist of 1–255 characters and contain letters, numbers, and symbols |
| Scenario type | Type of the scenario | - Event; |
| - Incident | ||
| Event Type | Appears only after selecting «events» as the scenario type | - Fire – fire detection; |
| - Animal – animal detection; | ||
| - Pedestrian – pedestrian detection; | ||
| - Vehicle – vehicle detection | ||
| Incident type | Appears only after selecting «incidents» as the scenario type | - Informational – an incident that does not require further review; |
| - Warning – an incident requiring separate review; | ||
| Status | Scenario activity status | - Enabled; |
| - Disabled | ||
| Handlers | ||
| Handler | The handler determines how the frame will be processed and saved as an event or incident | Handler configuration is described in Handlers section |
| Related Lists | ||
| List | The list with which the detected vehicles will be compared. If all attributes of the list entry match the attributes of the vehicles and LP, an event or incident is created. To exclude the recognition of vehicles and LP already in the list, activate the «Except» switch | Lists created by the Administrator. For more information, see Lists section |
To save the scenario, click the «Save» button in the top right corner.
Deleting a Scenario#
To delete a scenario, click on 
in the scenario row. A warning will appear (Figure 45), where you need to confirm the operation by clicking the «Delete» button.
Handlers#
The following handlers are available in CARS_Analytics:
| Handler | Description |
|---|---|
| Entering the region | Recognition of vehicle and LP attributes when entering the zone |
| Checkpoint | Recognition of vehicle and LP attributes when stopping in the zone |
| Move between zone | Recognition of vehicle and LP attributes when moving between specified zones |
| A pedestrian entering the region | Detection of a pedestrian when entering a specified zone |
| A animal entering the region | Detection of an animal when entering a specified zone |
| Counting the maximum number of animals | Counting the number of animales in the zone relative to the previous best shot (e.g., when a pack of dogs gathers) |
| Speed calculation | Speed calculation of the vehicle between specified zones |
| ANPR camera | Recognition of vehicle and LP attributes using an ANPR camera |
| License plate entering the zone | Recognition of LP attributes when entering the zone |
| Vehicles count | Counting the number of vehicles in the specified zone |
| Counting vehicles | Counting the number of vehicles in the zone relative to the previous best shot (e.g., parking space occupancy) |
| Smoke and fire detection | Fire detection in a specified zone |
| Frame based handler | Processing frames from the camera through a specified interval of 1 frame every X milliseconds (e.g., passing through a barrier gate) |
| Zone flow | Tracking the movement of vehicles in specified zones |
| Sabotage plugin | Detection of changes in the observation zone: lens shift and obstruction |
Notifications for deprecated handlers
The system provides notifications to inform about deprecated handlers. If a handler is blocked, its creation becomes impossible. In cases where the handler is no longer supported or requires replacement, an exclamation mark icon will appear next to it, and hovering over the icon will display a special notification (Figure 46).
Possible notifications that may appear next to the handler:
- «Deprecated handlers detected» — yellow exclamation mark icon. This notification informs that there are deprecated handlers in the system that require review or replacement.
- «No longer supported, please replace with the new version» — yellow exclamation mark icon. This notification informs that the handler is no longer supported and needs to be replaced with an updated version to ensure the system works correctly.
- «No longer supported, please remove» — red exclamation mark icon. This notification indicates that the handler has completely lost support and must be removed, as further use is not possible.
Entering the region#
The «Entering the region» handler is needed to create an event or incident when a vehicle crosses the area for crossing with a specified threshold.
The configuration of the «Entering the region» handler is done in the handler setup window after the handler type is set to «Entering the region» (Figure 47).
Handler parameters are shown in Table 12.
Table 14. Parameters of the «Entering the Area» handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Intersection policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area | ||
| Use unlimited bestshot count | Enabling the feature to send an unlimited number of bestshots (The parameters «Count of bestshots» and «Max count of bestshots» will not be considered, and the value in the «Max bestshot send periodicity» field must be greater than 0) | On/Off |
| Count of bestshots | The maximum number of best shots (bestshots) for a vehicle with a LP that can be issued for a vehicle track. The more bestshots are collected, the more accurate the information about the vehicle's attributes will be. Increasing the number of bestshots leads to higher CPU utilization | 1…N |
| - 1 – high probability of incorrect recognition results | ||
| - N – the time to select the best frame increases, but the quality is higher | ||
| Max count of bestshots | Count of bestshots – this is the maximum number of vehicles with a LP. Max count of bestshots – this is the maximum number of vehicles with and without a LP for a single track. Max count of bestshots - (minus) Count of bestshots = Max count of bestshots for vehicles without LP. The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher | ||
| Max delay first bestshot | The pause between the start of the track and the sending of the first frame to CARS_Stream to determine the best frame. The pause should be increased if the vehicle may not be in an optimal angle during the first few seconds of the track | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Max bestshot send periodicity | The frequency of sending the best frames for the track. The pause should be increased to ensure the vehicle is detected from different angles in the frames. A set of best frames from the track increases the probability of correct attribute recognition for the vehicle | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Delay for enable sending bestshot by periodicity | Pause before switching to the periodic best shot sending mode. During this pause, the best shot is searched for each frame; after the pause, the frames are sent at intervals specified in «Max bestshot send periodicity». This parameter applies only if «Max bestshot send periodicity» > 0. | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Use license plate area | Use an additional recognition zone for working with the part of the road where the LP is most visible | On/Off |
| Recommended to use: | ||
| - If there is a section of the road where the vehicle slows down and the number is less blurred; | ||
| - If the camera is installed in such a way that the LP is not visible in part of the track inside the recognition zone and it is necessary to specify the location of LP recognition | ||
| License plate area | Dropdown list to select the license plate recognition zone. The system recognizes the license plate only if the BBox of the license plate falls within this zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| License plate area threshold | The threshold for the intersection of the zone and the license plate BBox, above which the system considers the license plate to be in the zone | 0.01…1.00 |
| Use the vehicle key points (position) detection area | Enables the use of the position tracking area for the vehicle | On/Off |
| Vehicle key points (position) area | Dropdown list to select the vehicle position zone. The vehicle will be registered only if its position crosses the selected zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Vehicle key points (position) determination policy | Choose the algorithm for evaluating the vehicle’s position and checking if it is within the zone | - Periodic vehicle detection in the region — periodic evaluation of the vehicle's orientation in the zone |
| - First vehicle detection in the region — evaluation of the vehicle's orientation only for the first detection | ||
| Vehicle key points (position) crossing threshold | Threshold for crossing the vehicle’s position with the selected zone. The percentage of the vehicle’s position falling within the zone is calculated. If the calculated percentage exceeds the threshold, it is considered that the vehicle is in the zone, and it is then registered | 0.01…1.00 |
| Intersection area comparison policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area |
The results of the «Entering the regionc handler in the selected area are displayed in the event or incident card (Figure 48).
Usage goals with necessary handler parameters are shown in Table 15.
Table 15. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Logging of all vehicles passing on the road or by lanes (search, analysis, accounting, etc.) | vehicle | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection hould be selected so that the vehicle is fully visible without any overlaps | ||
| LP | - camera | |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the LP is fully visible, clear, and without overlaps | ||
| Traffic violation fixation – driving on the dedicated lane, driving onto the sidewalk, entering a restricted zone, etc | Vehicle in restricted zone and LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - full frames | ||
| Note: Camera should be positioned so that the restricted zone is in the frame |
Examples of using the handler with specified Region for intersection are provided in Table 16.
Table 16. Handler Usage Examples
| Description | Image |
|---|---|
| Logging of vehicles passing |  |
| Counting vehicles by lanes. 1 lane – 1 handler. |  |
| Fixing traffic violations – driving on the dedicated lane, driving onto the sidewalk, entering a restricted zone, etc. |  |
Example of using the track area and LP area (Figure 49).
| Element | Description |
|---|---|
| BBox | The vehicle's BBox and LP, whose coordinates and size will be compared with the areas |
 |
Region for intersection |
 |
Track area |
 |
LP area |
In this example, the vehicle will be recorded only after passing the barrier in the form of a road sign to avoid overlapping the LP with the sign’s post. The best LP frame will only be created when the LP is within the LP area.
Checkpoint#
The «Checkpoint» handler is necessary for creating an event or incident when a vehicle stops in a designated area, such as a parking lot or inspection zone.
The handler parameters for the «Checkpoint» handler are configured in the handler setup window after the handler type is set to «Checkpoint» (Figure 50).
It is not recommended to use this handler in places where movement may be obstructed.
Detection and tracking of vehicles and LP occur when entering the «Region for intersection». Best frame formation occurs only when stopping in the «Stop Area».
A vehicle is considered to have stopped when its position remains unchanged for the specified number of frames, which is set in the «Number of frames» field.
Example of vehicle recognition when using the «Checkpoint» handler (Figure 51).
| Element | Description |
|---|---|
| BBox | The vehicle's BBox, whose coordinates and size will be compared with the Region for intersection |
|
Region for intersection |
|
Stop area |
In this example, only vehicles that remain stationary for the specified number of frames in the «Stop Area» will be recorded. The best frames will be selected from the detections in the «Region for intersection».
Usage goals with the necessary handler parameters are shown in Table 17.
Table 17. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Control of public transport and traffic violation fixation | vehicle | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps | ||
| LP | - camera | |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the LP is fully visible, clear, and without overlaps | ||
| Traffic violation fixation – driving on the dedicated lane, driving onto the sidewalk, entering a restricted zone, etc | Vehicle in restricted zone and LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - full frames | ||
| Note: Camera should be positioned so that the restricted zone is in the frame |
Examples of using the handler with specified Region for intersection are provided in Table 16.
Table 18. Handler Usage Examples
| Description | Image |
|---|---|
| Control of public transport and traffic violation fixation – stopping in a forbidden place, controlling a bus stop, etc. |  |
Handler parameters are shown in Table 17.
Table 19. Parameters of the «Checkpoint» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Track end aggregation | Additional check for detections entering the zone at the end of the track if no detections met the handler condition when receiving the best shot | On/Off |
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Intersection policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area | ||
| Count of bestshots | Minimum number of frames for the current track to choose the best frame. If the track has fewer frames than the specified minimum in this parameter, the best frame will be selected, but the probability of correct attribute and LP symbol recognition will be lower. The more frames collected for the best shot, the more accurate the vehicle attributes will be recognized. Increasing the number of frames may reduce performance | 1…100 |
| - 1 – high probability of incorrect recognition results | ||
| - 100 – more time to select the best frame, but higher quality | ||
| Max count of bestshots | Count of bestshots – this is the maximum number of vehicles with a LP. Max count of bestshots – this is the maximum number of vehicles with and without a LP for a single track. Max count of bestshots - (minus) Count of bestshots = Max count of bestshots for vehicles without LP. The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher | ||
| Max delay first bestshot | The pause between the start of the track and the sending of the first frame to CARS_Stream to determine the best frame. The pause should be increased if the vehicle may not be in an optimal angle during the first few seconds of the track | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Max bestshot send periodicity | The frequency of sending the best frames for the track. The pause should be increased to ensure the vehicle is detected from different angles in the frames. A set of best frames from the track increases the probability of correct attribute recognition for the vehicle | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Detection Threshold | Minimum number of vehicle detections within one track for the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – disable the parameter | ||
| - 1 – Register events when receiving 1 detection. High probability of false events | ||
| - 10 – Register events with a minimum of 10 detections. High probability of missing events | ||
| Parameters threshold | Minimum number of recognized vehicle attributes (brand, model, belonging to an emergency service, special transport, public transport, vehicle type, country, color) on one best shot to trigger the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – all events/incidents are triggered, high probability of false events | ||
| - 10 – event/incident is triggered only when all vehicle attributes are recognized. High probability of missing events | ||
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Check the direction of movement | Boolean value to consider the direction of movement. When disabled, the direction of movement of the vehicle is not considered | On/Off |
| Angle of direction of movement | The movement direction angle of the vehicle relative to the frame in polar coordinates. Used to separate traffic flows, as the range of vehicle movement directions relative to a specified angle is +/- 90 degrees | 0…360 |
| - 0 – vehicle movement from left to right | ||
| - 90 – from bottom to top | ||
| - 180 – from right to left | ||
| - 270 – from top to bottom | ||
| Stop Area | Dropdown list to choose the recognition zone where the vehicle stop will be checked | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Number of frames | Minimum number of track frames in the track where the vehicle is stationary to record the fact of the vehicle stop | 1…240 |
| - 1 – The vehicle must be motionless for 1 second | ||
| - 240 – The vehicle must be motionless for 20 seconds | ||
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Use license plate area | Use an additional recognition zone for working with the part of the road where the LP is most visible. Recommended to use: | On/Off |
| - If there is a section of the road where the vehicle slows down and the number is less blurred; | ||
| - If the camera is installed in such a way that the LP is not visible in part of the track inside the recognition zone and it is necessary to specify the location of LP recognition | ||
| License plate area | Dropdown list to select the license plate recognition zone. The system recognizes the license plate only if the BBox of the license plate falls within this zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| License plate area threshold | The threshold for the intersection of the zone and the license plate BBox, above which the system considers the license plate to be in the zone | 0.01…1.00 |
Polar Coordinates of the Frame (Example of a vehicle moving at an angle of 200°) (Figure 52).
Move between zones#
The «Move between zones» handler is necessary to create an event or incident when a vehicle passes through specified zones, such as when moving through an intersection or passing through a barrier gate.
The handler parameters for the «Move between zones» handler are configured in the handler setup window after the handler type is set to «Move between zones» (Figure 53).
The handler is used to fix vehicles crossing zones in a specified sequence. It can be used, for example, for filtering by direction of movement – opposing or same direction.
Detection and tracking of vehicles and LP occur when entering the «Region for intersection». Event registration occurs only when the vehicle sequentially crosses the «First checkpoint» and the «Second checkpoint» within a single track.
Example of the «Move between zones» handler in operation (Figure 54).
| Element | Description |
|---|---|
| BBox | The vehicle's BBox, whose coordinates and size will be compared with the Region for intersection |
|
Region for intersection |
|
First checkpoint |
|
Second checkpoint |
In this example, an event will be created only for vehicles that sequentially pass through the first and second checkpoints, i.e., moving towards the camera.
Usage goals with the necessary handler parameters are shown in Table 20.
Table 20. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Determining the direction of movement – same direction, opposite direction, turn, U-turn | Vehicle and LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - first checkpoint | ||
| - second checkpoint | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps | ||
| Barrier Gate Opening | LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - first checkpoint | ||
| - second checkpoint | ||
| Note: The camera should be installed so that the barrier gate does not obscure zones in the frame |
Examples of using the handler with specified Region for intersection and intersection with checkpoints are provided in 21.
Table 21. Handler Usage Examples
| Description | Image |
|---|---|
| Control of vehicle movement at intersections, etc. |  |
| Barrier gate opening only for approaching vehicles |  |
Handler parameters are shown in Table 22.
Table 22. Parameters of the «Move between zones» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Track end aggregation | Additional check for detections entering the zone at the end of the track if no detections met the handler condition when receiving the best shot | On/Off |
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Intersection policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area | ||
| Use unlimited bestshot count | Enabling the feature to send an unlimited number of bestshots (The parameters «Count of bestshots» and «Max count of bestshots» will not be considered, and the value in the «Max bestshot send periodicity» field must be greater than 0) | On/Off |
| Count of bestshots | The maximum number of best shots (bestshots) for a vehicle with a LP that can be issued for a vehicle track. The more bestshots are collected, the more accurate the information about the vehicle's attributes will be. Increasing the number of bestshots leads to higher CPU utilization | 1…N |
| - 1 – high probability of incorrect recognition results | ||
| - N – the time to select the best frame increases, but the quality is higher | ||
| Count of bestshots | Minimum number of frames for the current track to choose the best frame. If the track has fewer frames than the specified minimum in this parameter, the best frame will be selected, but the probability of correct attribute and LP symbol recognition will be lower. The more frames collected for the best shot, the more accurate the vehicle attributes will be recognized. Increasing the number of frames may reduce performance | 1…100 |
| - 1 – high probability of incorrect recognition results | ||
| - 100 – more time to select the best frame, but higher quality | ||
| Max count of bestshots | Count of bestshots – this is the maximum number of vehicles with a LP. Max count of bestshots – this is the maximum number of vehicles with and without a LP for a single track. Max count of bestshots - (minus) Count of bestshots = Max count of bestshots for vehicles without LP. The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher | ||
| Max delay first bestshot | The pause between the start of the track and the sending of the first frame to CARS_Stream to determine the best frame. The pause should be increased if the vehicle may not be in an optimal angle during the first few seconds of the track | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Max bestshot send periodicity | The frequency of sending the best frames for the track. The pause should be increased to ensure the vehicle is detected from different angles in the frames. A set of best frames from the track increases the probability of correct attribute recognition for the vehicle | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Detection Threshold | Minimum number of vehicle detections within one track for the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – disable the parameter | ||
| - 1 – Register events when receiving 1 detection. High probability of false events | ||
| - 10 – Register events with a minimum of 10 detections. High probability of missing events | ||
| Parameters threshold | Minimum number of recognized vehicle attributes (brand, model, belonging to an emergency service, special transport, public transport, vehicle type, country, color) on one best shot to trigger the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – all events/incidents are triggered, high probability of false events | ||
| - 10 – event/incident is triggered only when all vehicle attributes are recognized. High probability of missing events | ||
| First checkpoint | Dropdown list to select the zone where the vehicle’s BBox crossing is checked for marking the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Intersection with the first checkpoint | Minimum percentage of crossing detection with the «First checkpoint», required to form an event/incident. When the threshold is exceeded, the crossing will be recorded. If the value is less than 0 or greater than 100, an error will occur. A warning with the two closest valid values will appear if more than three decimal places are entered | 0.01…100 |
| Second checkpoint | Dropdown list to select the zone where the vehicle’s BBox crossing is checked for marking the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Intersection with the second checkpoint | Minimum percentage of crossing detection with the «Second checkpoint», required to form an event/incident. When the threshold is exceeded, the crossing will be recorded. If the value is less than 0 or greater than 100, an error will occur. A warning with the two closest valid values will appear if more than three decimal places are entered | 0.01…100 |
| Min count of intersections with the checkpoint | Minimum number of crossings with the «Second checkpoint», required to form an event. Increasing the value helps reduce false events | 1…48 |
| Only first then second | Best shots are sent when the vehicle is detected in the first checkpoint, if it does not appear in the second, and only after appearing in the second checkpoint, regardless of its presence in the first. This mode is suitable for large objects such as buses or trucks. Only one rule can be activated at a time | On/Off |
| First then only second | Best shots are sent when the vehicle is detected in the first checkpoint, regardless of its appearance in the second checkpoint, and then only when it appears in the second checkpoint. This mode is suitable for large objects such as buses or trucks. Only one rule can be activated at a time | On/Off |
| Only first then only second | Best shots are sent only when the vehicle crosses the first checkpoint separately, and then the second. The zones must be arranged so that the vehicle is not in both checkpoint at the same time. Only one rule can be activated at a time | On/Off |
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Use license plate area | Use an additional recognition zone for working with the part of the road where the LP is most visible | On/Off |
| Recommended to use: | ||
| - If there is a section of the road where the vehicle slows down and the number is less blurred; | ||
| - If the camera is installed in such a way that the LP is not visible in part of the track inside the recognition zone and it is necessary to specify the location of LP recognition | ||
| License plate area | Dropdown list to select the license plate recognition zone. The system recognizes the license plate only if the BBox of the license plate falls within this zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| License plate area threshold | The threshold for the intersection of the zone and the license plate BBox, above which the system considers the license plate to be in the zone | 0.01…1.00 |
Figures 55–57 show examples describing the operation of the rules:
1․ Only first then second (Figure 55);
In the first frame, the vehicle enters the first checkpoint (Zone 1), and in the second frame, it enters the second checkpoint (Zone 2), followed by the formation of the best frame for the vehicle and LP.
2․ First then only second (Figure 56);
3․ Only first then only second (Figure 57);
A pedestrian entering the region#
The «A pedestrian entering» handler is necessary for creating an event or incident when a person crosses the area for crossing with a specified threshold.
The configuration of the «A pedestrian entering» handler is done in the handler setup window after the handler type is set to «A pedestrian entering» (Figure 58).
Example of pedestrian recognition (Figure 59).
| Element | Description | Intersection square with Region for intersection |
|---|---|---|
| BBox | The pedestrian's BBox, whose coordinates and size will be compared with the Region for intersection | - |
|
Region for intersection | - |
 |
Percentage of the Region for intersection covered by BBox 1 | 0,6 |
Usage goals with necessary handler parameters are shown in Table 23.
Table 23. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Control of people in restricted areas | Pedestrian | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the pedestrian is fully visible without overlaps | ||
| Analysis and accounting human stream crossing specified area | Pedestrian | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the pedestrian is fully visible without overlaps |
Examples of using the handler with specified Region for intersection are provided in Table 24.
Table 24. Handler Usage Examples
| Description | Image |
|---|---|
| Control of people in restricted areas |  |
| Analysis and accounting human stream crossing specified area |  |
Handler parameters are shown in Table 25.
Table 25. Parameters of the «A pedestrian entering» handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a pedestrian enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and pedestrian detection, above which the system considers the pedestrian to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Count of bestshots | The maximum number of best shots (bestshots) for a pedestrian that can be issued for a pedestrian track. Increasing the number of bestshots leads to higher CPU utilization | 1…N |
| - 1 – high probability of incorrect recognition results | ||
| - N – the time to select the best frame increases, but the quality is higher | ||
| Max count of bestshots | The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher |
A animal entering the region#
The «A animal entering» handler is necessary for creating an event or incident when a person crosses the area for crossing with a specified threshold.
The configuration of the «A animal entering» handler is done in the handler setup window after the handler type is set to «A animal entering» (Figure 60).
Example of animal recognition (Figure 61).
| Element | Description | Intersection square with Region for intersection |
|---|---|---|
| BBox | The animal's BBox, whose coordinates and size will be compared with the Region for intersection | - |
|
Region for intersection | - |
|
Percentage of the Region for intersection covered by BBox 1 | 0.5 |
Usage goals with necessary handler parameters are shown in Table 26.
Table 26. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Control of animals in restricted areas | Animal | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the animal is fully visible without overlaps |
Examples of using the handler with specified Region for intersection are provided in Table 25.
Table 27. Handler Usage Examples
| Description | Image |
|---|---|
| Control of animals in restricted areas | |
Handler parameters are shown in Table Table 26.
Table 28. Parameters of the «A animal entering» handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when an animal enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and animal detection, above which the system considers the animal to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Count of bestshots | The maximum number of best shots (bestshots) for a animal can be issued for a animal track. Increasing the number of bestshots leads to higher CPU utilization | 1…N |
| - 1 – high probability of incorrect recognition results | ||
| - N – the time to select the best frame increases, but the quality is higher | ||
| Max count of bestshots | The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher |
Counting the maximum number of animals#
The «Counting the maximum number of animals» handler is designed to register an event or incident when the number of animals in a specified area changes.
For this handler, the current number of animals is displayed in the event and incident tables, for example, «Animals count: 4».
The configuration of the «Counting the maximum number of animals» handler parameters is done in the handler configuration window after the handler type is set to «Counting the maximum number of animals» (Figure 62).
Handler parameters are shown in Table Table 36.
Table 29. Parameters of the «Counting the maximum number of animals» handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone where an animal's entry will trigger the best frame to be issued | Created zones for the selected camera. The zone creation process is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and animal detection, above which the system considers the animal to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Detections count threshold | Threshold value for the number of animals required to register the best shot. The value must be greater than or equal to 1 | 1…100 |
The results of counting the maximum number of animals in the selected observation area are displayed in the event or incident card (Figure 63).
Table 30. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Tracking the formation of an animal group in the zone | Animals | - camera; |
| - region for intersection; | ||
| - detections count threshold | ||
| Note: The event will only be generated when the number of animals in the observation area changes |
Speed calculation#
The «Speed Calculation» handler works similarly to the «Entering the region» handler, but additionally allows the recording of vehicle speed.
The parameters for the «Speed Calculation» handler are configured in the handler setup window after the handler type is set to «Speed Calculation» (Figure 64).
The event or incident is created when the vehicle crosses the crossing area, and speed is calculated and registered after the vehicle sequentially crosses the Start and Finish regions.
Detection and tracking of vehicles and LP occur when entering the Region for intersection. When determining the vehicle's BBox, the center point of the BBox is calculated and used to check the intersection with the Start and Finish regions.
It is recommended to form the Start and Finish regions with sufficient width for proper handler operation.
Speed calculation occurs after the vehicle's BBox exits the final area at the end of the track.
Speed calculation groups are created for the handler, where individual parameters are set (Figure 65). Within one handler, multiple speed calculation groups can be created for a single video stream (e.g., groups for same-direction and opposite-direction traffic, etc.).
Example of vehicle recognition when using the «Speed Calculation» handler (Figure 66).
| Element | Description |
|---|---|
| BBox | The vehicle's BBox, whose coordinates and size will be compared with the Region for intersection |
|
Region for intersection |
|
Start region |
 |
Finish region |
Usage goals with the necessary handler parameters are shown in Table 31.
Table 31. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Monitoring vehicle speeds in a specified area (data collection, speed analysis, etc.) | vehicle | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - start region | ||
| - finish region | ||
| - distance | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps. Start and Finish regions have sufficient width to intersect with geometric midpoint of vehicle BBox |
Examples of using the handler with specified Region for intersection are provided in Table 32.
Table 32. Handler Usage Examples
| Description | Image |
|---|---|
| Monitoring vehicle speeds in a specified area (data collection, speed analysis, etc.) |  |
Handler parameters are shown in Table 33.
Table 33. Parameters of the «Speed Calculation» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Intersection policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area | ||
| Count of bestshots | Minimum number of frames for the current track to choose the best frame. If the track has fewer frames than the specified minimum in this parameter, the best frame will be selected, but the probability of correct attribute and LP symbol recognition will be lower. The more frames collected for the best shot, the more accurate the vehicle attributes will be recognized. Increasing the number of frames may reduce performance | 1…100 |
| - 1 – high probability of incorrect recognition results | ||
| - 100 – more time to select the best frame, but higher quality | ||
| Max count of bestshots | Count of bestshots – this is the maximum number of vehicles with a LP. Max count of bestshots – this is the maximum number of vehicles with and without a LP for a single track. Max count of bestshots - (minus) Count of bestshots = Max count of bestshots for vehicles without LP. The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher | ||
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Use license plate area | Use an additional recognition zone for working with the part of the road where the LP is most visible. Recommended to use: | On/Off |
| - If there is a section of the road where the vehicle slows down and the number is less blurred; | ||
| - If the camera is installed in such a way that the LP is not visible in part of the track inside the recognition zone and it is necessary to specify the location of LP recognition | ||
| License plate area | Dropdown list to select the license plate recognition zone. The system recognizes the license plate only if the BBox of the license plate falls within this zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| License plate area threshold | The threshold for the intersection of the zone and the license plate BBox, above which the system considers the license plate to be in the zone | 0.01…1.00 |
| Speed Calculation group | ||
| Line title | Title of the speed calculation group | - |
| Start region | Dropdown list to select the zone where the vehicle intersections are checked for fixing the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Finish region | Dropdown list to select the zone where the vehicle intersections are checked for fixing the final control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Distance | Distance between the Start and End regions in meters | 0.01…1000 |
| Track area | Dropdown list to select the detection zone where the vehicle intersections are checked. The introduction of an additional zone will allow you to cut off events that do not need to be recorded | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Track area threshold | The relative area of the «Recognition Zone» overlapped by the BBox of the vehicle, when exceeding the threshold of which the vehicle will be taken into account to form the best frames. The higher the threshold, the more accurate the detection of the vehicle and LP will be, but the number of passes will also increase | 0.01…1.00 |
The speed calculation results are displayed in the event or incident card (Figure 67).
ANPR Camera#
The «ANPR Camera» handler allows for registering objects detected by the ANPR camera.
The handler parameters for the «ANPR Camera» handler are configured in the handler setup window after the handler type is set to «ANPR Camera» (Figure 68).
Detection zone and recognition zone settings for the «ANPR Camera» handler are available in the Cameras section under the «Settings» block for the ANPR camera source type.
Handler parameters are shown in Table 34.
Table 34. Parameters of the «ANPR Camera» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
Licence plate entering the zone#
The «LP entering the zone» handler is necessary for creating an event or incident when the LP crosses the crossing area with a specified threshold.
The parameters for the «LP entering the zone» handler are configured in the handler setup window after the handler type is set to «LP entering the zone» (Figure 69).
Handler parameters are shown in Table 35.
Table 35. Parameters of the «LP entering the zone» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Bestshot send periodicity | Frequency of sending best shots | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – in milliseconds | ||
| Event uniqueness time | Maximum duration at which two identical LP received will be counted for the same event | 0…60000 |
Example of vehicle recognition when using the «LP entering the zone» handler (Figure 70).
Usage goals with the necessary handler parameters are shown in Table 36.
Table 36. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| License plate fixation in a section with a stop or barrier checkpoint to prioritize license plate events | LP | - camera; |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: The area to be crossed shall be selected so that the license plate can be seen in its entirety, clearly and without overlap |
Vehicles count#
The «Vehicles count» handler is necessary for creating an event or incident when counting the number of vehicles in a specified region with the specified threshold.
The parameters for the «Vehicles count» handler are configured in the handler setup window after the handler type is set to «Vehicles count» (Figure 71).
Handler parameters are shown in Table 37.
Table 37. Parameters of the «Vehicles count» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Monitoring region | Dropdown list to choose the zone where the vehicles will be counted for the scenario | Created zones for the selected camera. The process of creating recognition zones is described in Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Number of vehicles | Minimum number of vehicles for which the event is generated | 0…10000 |
| Bestshot send periodicity | Frequency of sending best shots | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – in milliseconds |
The results of vehicle counting in the selected observation area are displayed in the event or incident card (Figure 72).
In case of event registration with exceeding the parameter «Number of vehicles» in the control zone, the next event is registered only if the number of vehicles in the control zone increases. An example of the «Vehicles count» handler operation is given in Table 38.
Table 38. Handler Usage Examples
| Frame processing stage | Event description | Event status |
|---|---|---|
| (1) | There are more than X number of cars in the control zone | Registered |
| (2) | At a defined interval, when counting vehicles in the control zone, X vehicles | Not registered |
| (3) | At a defined interval, when counting vehicles in the control zone, X - 1 vehicles | Not registered |
| (4) | At a defined interval, when counting the vehicles in the control zone, X vehicles | Registered |
| (5) | At a defined interval, when counting vehicles in the control zone, X + 1 vehicles | Registered |
Usage goals with the necessary handler parameters are shown in Table 39.
Table 39. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Detection of the number of vehicles in the control zone, without taking into account other attributes. | Vehicle | - camera; |
| - monitoring region | ||
| - min threshold intersection | ||
| - number of vehicles | ||
| Note: The event will be generated only when the specified value of the number of vehicles in the monitoring area is reached |
Counting vehicles#
The «Counting vehicles» handler is designed to count the number of vehicles in a specified zone. The handler counts the vehicles in the zone and generates an event whenever there is a change. It is suitable for assessing the occupancy of one or more parking spaces. In the event and incident tables, the current number of vehicles and the change in number are displayed, for example, «Number of vehicles: 2 −1».
The parameters for the «Counting vehicles» handler are configured in the handler setup window after the handler type is set to «Counting vehicles» (Figure 73).
Handler parameters are shown in Table 40.
Table 40. Parameters of the «Counting vehicles» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Monitoring region | Dropdown list to choose the zone where the vehicles will be counted for the scenario | Created zones for the selected camera. The process of creating recognition zones is described in Cameras section |
| Bestshot send periodicity | Frequency of sending best shots | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Vehicle key points (position) crossing threshold | Threshold for the vehicle’s position crossing the specified zone. The percentage of the vehicle's position in the zone is calculated. If the calculated percentage exceeds the threshold, the vehicle is considered to be in the zone | 0.01…1.00 |
| Intersection area comparison policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area |
The results of the «Counting vehicles» in the selected observation area are displayed in the event or incident card (Figure 74).
When clicking on any frame, detailed information about the frame will open (Figure 75).
A pop-up window appears containing:
- Frame ID – the unique identifier of the frame
- Time – the date and time when the event was detected
- Type – the type of the frame
- Label – the name of the frame type
- Recognition zones – using checkboxes, you can display the recognition zones on the frame
- Detected vehicles – by enabling the checkboxes, you can highlight detected objects in the frame using BBox (borders of the vehicle) and the vehicle position (projection of the vehicle on the road)
Table 41. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Counting vehicles in a parking zone, without considering other attributes | Vehicle | - camera; |
| - monitoring region | ||
| - bestshot send periodicity | ||
| - vehicle key points (position) crossing threshold | ||
| - intersection area comparison policy | ||
| Note: An event will be generated only when the number of vehicles in the observation area changes |
Smoke and fire detection#
The «Smoke and Fire Detection» handler is used to create an event or incident when smoke and/or fire is detected in a specified region, based on a set threshold.
The handler parameters for «Smoke and Fire Detection» are configured in the handler settings window after the handler type is set to «Smoke and Fire Detection» (Figure 76).
The handler parameters are shown in Table 42.
Table 42. Handler parameters
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Monitoring region | Dropdown list for selecting the zone with which smoke and fire are checked for the scenario to work | Created zones for the selected camera. The process of creating recognition zones is described in Cameras section |
| Bestshot send periodicity | Frequency of sending best shots | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Min threshold intersection | The threshold for zone intersection and smoke and fire detection, above which the system considers the smoke and fire to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
The results of smoke and fire detection in the selected monitoring area are displayed in the event or incident card (Figure 77).
Usage goals with the necessary handler parameters are shown in Table 43.
Table 43. Usage Goals
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Logging of Warning events in case of smoke and/or fire detection | - Fire | - camera; |
| - Black smoke | - monitoring region | |
| - White smoke | - minimal threshold of intersection | |
| Note: The input image is divided into 6 zones, each of which detects one of the incident attributes |
Frame based handler#
The «Frame based handler» is necessary for processing the video stream when vehicles are moving at low speeds (e.g., passing through a barrier). This handler processes frames of the video stream at a specified interval, reducing the system load while maintaining license plate recognition accuracy in the control zone (e.g., 1 frame every X milliseconds).
The handler parameters are configured in the handler settings window after the handler type is set to «Frame based handler» (Figure 78).
Handler parameters are shown in Table 44.
Table 44. Parameters of the «Frame based handler» handler
| Name | Description | Possible Values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Bestshot send periodicity | Frequency of sending best shots | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Life time track without updates | The time during which the track is retained without updates via best shots. This value indicates how many milliseconds the track can stay without updates before being deleted. Increasing this time can be helpful if the vehicle's track is temporarily blocked by an external object | 0…10000 |
| - 0 – track is deleted immediately after no new best shot is received | ||
| - 10000 – track is retained for 10 seconds without updates | ||
| Acceptable orientations | Dropdown list for selecting the allowed vehicle orientation. The vehicle will be registered if the orientation matches one of the selected values Forward, Backward, Forward Left, Forward Right, Backward Left, Backward Right (multiple selections allowed) | |
| Min threshold descriptor match score | For combining best shots of a vehicle into one track, matching by descriptor is used. Two best shots of a vehicle are considered part of one track if the minimum descriptor similarity threshold is passed during matching | 0.01…1.00 |
| - 0.01 – minimum similarity, more matches but higher risk of errors | ||
| - 1.00 – strict similarity check, minimal error probability, but multiple tracks may be created for one vehicle |
The results of the «Frame based handler» in the selected observation area are displayed in the event or incident card (Figure 79).
Usage goals with the necessary handler parameters are shown in Table 45.
Table 45. Usage Goals
| Purpose of use | Important parameters for this purpose |
|---|---|
| Register Vehicle when passing through the barrier (necessary to detect: Vehicle) | - Camera; |
| - region for intersection | |
| - life time track without updates | |
| - acceptable orientations | |
| - min threshold descriptor match score | |
| Note: Event is recorded when the vehicle passes through the control zone |
Zone flow#
The «Zone flow» handler is designed to track the movement of vehicles within specified areas. This handler combines parameters from different handlers for more flexible configuration, allowing for effective use in scenarios related to vehicle tracking. It can be used to register each vehicle that enters the frame, as well as to monitor vehicles crossing defined zones. A key feature of the handler is the accounting for vehicle orientation, movement vector, and precise positioning, which allows the traffic to be divided into oncoming and same-direction traffic.
The configuration of the «Zone flow» handler is done in the handler settings window after the handler type is set to «Zone flow» (Figure 80).
The handler parameters are listed in Table 46.
Table 46. Parameters of the «Zone flow» Handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list containing handlers that process the detected information for this scenario | Available handlers (for more information, see beginning of the Handlers section) |
| Camera | Camera from which the video stream frames will be received for handler operation | Cameras added in the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
| Camera and Recognition Zone Settings | ||
| Region for intersection | Dropdown list to select the zone, where, when a vehicle enters, the best frame will be issued | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Min threshold intersection | The threshold for zone intersection and vehicle detection, above which the system considers the vehicle to be in the zone. When the threshold is exceeded, the system provides the best frame | 0.01…1.00 |
| Intersection policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area | ||
| Acceptable orientations | Dropdown list to select the acceptable vehicle orientations. The best frame will be issued for the vehicle if its orientation matches one of the selected values in the list | Forward, Backward, Forward Left, Forward Right, Backward Left, Backward Right (multiple options can be selected) |
| Vehicle orientation detection policy | The algorithm for evaluating the vehicle's orientation | - Periodic vehicle detection in the region — periodic evaluation of the vehicle's orientation in the zone |
| - First vehicle detection in the region — evaluation of the vehicle's orientation only for the first detection | ||
| Move between zones group | ||
| The group active | The value of the activity of these settings | On/Off |
| First checkpoint | Dropdown list to select the zone where the vehicle’s BBox crossing is checked for marking the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Intersection with the first checkpoint | Minimum percentage of crossing detection with the «First checkpoint», required to form an event/incident. When the threshold is exceeded, the crossing will be recorded. If the value is less than 0 or greater than 100, an error will occur. A warning with the two closest valid values will appear if more than three decimal places are entered | 0.01…100 |
| Second checkpoint | Dropdown list to select the zone where the vehicle’s BBox crossing is checked for marking the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Intersection with the second checkpoint | Minimum percentage of crossing detection with the «Second checkpoint», required to form an event/incident. When the threshold is exceeded, the crossing will be recorded. If the value is less than 0 or greater than 100, an error will occur. A warning with the two closest valid values will appear if more than three decimal places are entered | 0.01…100 |
| Min count of intersections with the checkpoint | Minimum number of crossings with the «Second checkpoint», required to form an event. Increasing the value helps reduce false events | 1…48 |
| Only first then second | Best shots are sent when the vehicle is detected in the first checkpoint, if it does not appear in the second, and only after appearing in the second checkpoint, regardless of its presence in the first. This mode is suitable for large objects such as buses or trucks. Only one rule can be activated at a time | On/Off |
| First then only second | Best shots are sent when the vehicle is detected in the first checkpoint, regardless of its appearance in the second checkpoint, and then only when it appears in the second checkpoint. This mode is suitable for large objects such as buses or trucks. Only one rule can be activated at a time | On/Off |
| Only first then only second | Best shots are sent only when the vehicle crosses the first checkpoint separately, and then the second. The zones must be arranged so that the vehicle is not in both checkpoint at the same time. Only one rule can be activated at a time | On/Off |
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Speed Calculation group | ||
| The group active | The value of the activity of these settings | On/Off |
| Line title | Title of the speed calculation group | - |
| Start region | Dropdown list to select the zone where the vehicle intersections are checked for fixing the initial control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Finish region | Dropdown list to select the zone where the vehicle intersections are checked for fixing the final control point | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Distance | Distance between the Start and End regions in meters | 0.01…1000 |
| Track area | Dropdown list to select the detection zone where the vehicle intersections are checked. The introduction of an additional zone will allow you to cut off events that do not need to be recorded | Created zones for the selected camera. The zone creation process is described in Cameras section |
| Track area threshold | The relative area of the «Recognition Zone» overlapped by the BBox of the vehicle, when exceeding the threshold of which the vehicle will be taken into account to form the best frames. The higher the threshold, the more accurate the detection of the vehicle and LP will be, but the number of passes will also increase | 0.01…1.00 |
| Extended params | ||
| Use unlimited bestshot count | Enabling the feature to send an unlimited number of bestshots (The parameters «Count of bestshots» and «Max count of bestshots» will not be considered, and the value in the «Max bestshot send periodicity» field must be greater than 0) | On/Off |
| Count of bestshots | The maximum number of best shots (bestshots) for a vehicle with a LP that can be issued for a vehicle track. The more bestshots are collected, the more accurate the information about the vehicle's attributes will be. Increasing the number of bestshots leads to higher CPU utilization | 1…N |
| - 1 – high probability of incorrect recognition results | ||
| - N – the time to select the best frame increases, but the quality is higher | ||
| Max count of bestshots | Count of bestshots – this is the maximum number of vehicles with a LP. Max count of bestshots – this is the maximum number of vehicles with and without a LP for a single track. Max count of bestshots - (minus) Count of bestshots = Max count of bestshots for vehicles without LP. The value of this parameter must be greater than or equal to the value of the «Count of bestshots» parameter | 1…10000 |
| - 1 – high probability of incorrect recognition results due to low quality of the best frame | ||
| - 10000 – the time to select the best frame increases, but the quality is higher | ||
| Max delay first bestshot | The pause between the start of the track and the sending of the first frame to CARS_Stream to determine the best frame. The pause should be increased if the vehicle may not be in an optimal angle during the first few seconds of the track | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Max bestshot send periodicity | The frequency of sending the best frames for the track. The pause should be increased to ensure the vehicle is detected from different angles in the frames. A set of best frames from the track increases the probability of correct attribute recognition for the vehicle | 0…60000 |
| - 0 – no delay | ||
| - 1…60000 – pause in milliseconds | ||
| Delay for enable sending bestshot by periodicity | Pause before switching to the periodic best shot sending mode. During this pause, the best shot is searched for each frame; after the pause, the frames are sent at intervals specified in «Max bestshot send periodicity». This parameter applies only if «Max bestshot send periodicity» > 0 | 0…10000 |
| - 0 – no delay | ||
| - 1…10000 – pause in milliseconds | ||
| Priority of Detection Selection | Algorithm for selecting the best detection and the best shot | - Min distance to the bottom – the best detection is the one closest to the bottom edge of the frame; |
| - Min detection area – the best detection is the one with the smallest area; | ||
| - Max detection area – the best detection is the one with the largest area; | ||
| - Min frame ID – the best detection is the one with the smallest frame ID, i.e., detection when the object «enters» the frame; | ||
| - Max frame ID – the best detection is the one with the largest frame ID, i.e., detection when the object «exits» the frame; | ||
| - Min intersection – the best detection is the one with the smallest intersection ratio with the crossing area; | ||
| - Max intersection – the best detection is the one with the largest intersection ratio with the crossing area | ||
| Parameters threshold | Minimum number of recognized vehicle attributes (brand, model, belonging to an emergency service, special transport, public transport, vehicle type, country, color) on one best shot to trigger the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – all events/incidents are triggered, high probability of false events | ||
| - 10 – event/incident is triggered only when all vehicle attributes are recognized. High probability of missing events | ||
| Detection Threshold | Minimum number of vehicle detections within one track for the scenario. Increase the threshold to reduce the number of false event triggers when the vehicle appears in the «Region for intersection» | 0…10 |
| - 0 – disable the parameter | ||
| - 1 – Register events when receiving 1 detection. High probability of false events | ||
| - 10 – Register events with a minimum of 10 detections. High probability of missing events | ||
| License plates zone params | ||
| The group active | The value of the activity of these settings | On/Off |
| License plate area | Dropdown list to select the license plate recognition zone. The system recognizes the license plate only if the BBox of the license plate falls within this zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| License plate area threshold | The threshold for the intersection of the zone and the license plate BBox, above which the system considers the license plate to be in the zone | 0.01…1.00 |
| Key points zone params | ||
| The group active | The value of the activity of these settings | On/Off |
| Vehicle key points (position) area | Dropdown list to select the vehicle position zone. The vehicle will be registered only if its position crosses the selected zone. | Created zones for the selected camera. The process of creating recognition zones is described in the Cameras section |
| Vehicle key points (position) determination policy | Choose the algorithm for evaluating the vehicle’s position and checking if it is within the zone | - Periodic vehicle detection in the region — periodic evaluation of the vehicle's orientation in the zone |
| - First vehicle detection in the region — evaluation of the vehicle's orientation only for the first detection | ||
| Vehicle key points (position) crossing threshold | Threshold for crossing the vehicle’s position with the selected zone. The percentage of the vehicle’s position falling within the zone is calculated. If the calculated percentage exceeds the threshold, it is considered that the vehicle is in the zone, and it is then registered | 0.01…1.00 |
| Intersection area comparison policy | Dropdown list that defines which area is used to calculate the percentage of vehicle crossing with the selected zone: zone (area of intersection/area of zone), vehicle detection (area of intersection/area of detection), or the smaller of the areas (area of intersection/min(zone area, detection area)) | - Using region area; |
| - Using detection area; | ||
| - Using smaller area |
The results of the handler's operation within the selected observation area are displayed in the event or incident card (Figure 81).
The goals of use, along with the necessary handler parameters, are provided in Table 47.
Table 47. Description of the goals for using the «Zone flow» handler
| Purpose of use | Necessary to detect… | Important parameters for this purpose |
|---|---|---|
| Control of public transport and traffic violation fixation | vehicle | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps | ||
| LP | - camera | |
| - region for intersection | ||
| - min threshold intersection | ||
| Determining the direction of movement – same direction, opposite direction, turn, U-turn | Vehicle and LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - first checkpoint | ||
| - second checkpoint | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps | ||
| Barrier Gate Opening | LP | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - first checkpoint | ||
| - second checkpoint | ||
| Note: The camera should be installed so that the barrier gate does not obscure zones in the frame | ||
| Monitoring vehicle speeds in a specified area (data collection, speed analysis, etc.) | vehicle | - camera |
| - region for intersection | ||
| - min threshold intersection | ||
| - start region | ||
| - finish region | ||
| - distance | ||
| Note: Region for intersection should be selected so that the vehicle is fully visible without overlaps. Start and Finish regions have sufficient width to intersect with geometric midpoint of vehicle BBox |
Sabotage plugin#
The «Sabotage plugin» handler is designed to detect camera angle changes, camera shifts, or partial/full obstruction of the camera. This allows for the timely detection of attempts to interfere with the camera system. The handler responds to frame changes and generates an incident whenever sabotage is detected. In the incident card, only the frame at the beginning and the end of the track are displayed.
Attention!
To make the «Sabotage plugin» handler work, several steps must be followed:
1․ Create a separate scenario with any scenario type. 2․ Add and configure any handler in the scenario. 3․ Create a new scenario with the type «Incidents» (Figure 82). 4․ Add the «Sabotage plugin» handler to the new scenario (Figure 83).
Both scenarios must be created for the same camera.
The handler parameters are listed in Table 48.
Table 48. Parameters of the «Sabotage plugin» handler
| Name | Description | Possible values |
|---|---|---|
| General | ||
| Type | Dropdown list with handlers responsible for detecting and aggregating information for the given scenario | Available handlers (more details in the Handlers section at the beginning) |
| Camera | Camera from which video stream frames will be obtained for the handler's operation | Cameras added to the system |
| Comment | User comment | - |
| Status | Handler activity status | On/Off |
When a frame change occurs, an alarm icon will appear next to the «Cameras» tab, indicating a problem (Figure 84).
Clicking on the exclamation mark icon will take you to the «Incidents» section, filtered by scenario and camera (Figure 85).
The results of the handler's operation are displayed in the incident card. Depending on the situation, the incident may be related to one of the types of sabotage.
An example of camera obstruction is shown below (Figure 86).
An example of camera shift is shown below (Figure 87).
Goals of use with necessary handler parameters are listed in Table 49.
Table 49. Usage Goals
| Purpose of use | Important parameters for this purpose |
|---|---|
| Detection of attempted physical intervention | - Camera; |
| Monitoring system integrity | - Camera; |
| Detecting changes in the detection area | - Camera; |