«Events» Section#
The «Events» section is designed to display detection events for vehicles, pedestrians, and animals, as well as recognition of vehicle attributes and LP.
The section is used to display all events from incoming video stream images. Receiving and displaying events is performed with minimal delays in a mode close to real time.
The general view of the «Events» section (Figure 10).
By default, the last 10 events are displayed on the page. To view previous events, navigate to other pages. Page navigation is done using the pagination buttons at the bottom of the page (Figure 11).
The number of events per page can be changed using the dropdown list next to the pagination buttons.
When navigating between event pages, new events will not be displayed in the table. This is to prevent the rows from «moving» during event table viewing. Once the user returns to the first page, all new events are automatically loaded.
The description of the event table content is provided in Table 3.
Table 3. Events table
| Name | Description |
|---|---|
| ID | Event ID |
| Camera | The name of the camera that recorded the event |
| Date | Date and time of event registration |
| Scenario | The scenario of the event registration |
| Event type | The type of the recorded event, set by the scenario |
| Information | |
| Vehicle Frame | Vehicle crop used for vehicle attribute recognition |
| Vehicle Type | Vehicle attribute showing the vehicle type |
| Brand | Vehicle attribute showing the recognized vehicle brand |
| Model | Vehicle attribute showing the recognized vehicle model |
| Color | Vehicle attribute showing the recognized vehicle color |
| Emergency services | Attribute of a vehicle indicating its affiliation with emergency services |
| LP Frame | LP crop used for LP attribute recognition |
| Number | LP attribute providing information about LP symbols |
| Vehicle Country | LP attribute indicating country of vehicle registration |
| Pedestrian Frame | Pedestrian crop (relevant for pedestrian detection scenarios) |
| Animal Frame | Animal crop (relevant for animal detection scenarios) |
| Fire type | Attribute indicating the type of fire detected in the monitoring area. Possible values are displayed in detection priority order: Fire > Black smoke > White smoke > No fire. (Relevant for fire detection scenarios) |
The numbers next to each attribute represent the recognition quality of that attribute. The rating is determined within the range from 0 to 1; the closer the value is to 1, the more accurate the recognition result is.
Event Card#
The event card shows extended information about the event. To access the event card, click on the event ID (Figure 12).
Example of an Event card (Figure 13).
The description of the information presented in the event card is provided in Table 4.
Table 4. Event card description
| Section | Name | Description |
|---|---|---|
| General | ID | Event ID |
| Registration Date, Time | Event registration date and time | |
| Camera | The name of the camera that recorded the event | |
| Scenario | The name of the scenario of the event registration. The scenario name is a link to the scenario settings window | |
| Scenario type | The type of scenario that detected the event | |
| Comment | A link that opens a comment entry window when clicked. After adding a comment, the link changes to the text of the comment | |
| Track ID | Track ID where the event was recorded | |
| Frame ID | Frame ID containing the object image | |
| Frames | Track Frames | Object images in original (maximum) resolution: |
| - Bestshot – the best frame of the track. | ||
| - Track start frame – the first frame where the object was detected | ||
| - Track end frame – the last frame where the object tracking ended | ||
| Attributes | Type | Shows the type of vehicle |
| Brand | Shows the recognized vehicle brand | |
| Model | Shows the recognized vehicle model | |
| Vehicle Orientation | Vehicle attribute indicating the vehicle's orientation in space | |
| Emergency services | Vehicle attribute indicating its affiliation with emergency services | |
| Public transport | Vehicle attribute indicating its affiliation with public transport | |
| Special transport | Vehicle attribute indicating its affiliation with special transport | |
| Color | Vehicle attribute showing the recognized vehicle color | |
| Number of axles | Vehicle attribute showing the recognized number of vehicle axles | |
| Fire types (Fire, Black Smoke, White Smoke, No Fire) | Attributes showing the types of fires detected and the number of zones with detections of the specified type. Available only when working with the «Smoke and Fire Detection» handler | |
| Frame | Object crop. Vehicle crop is used for attribute recognition | |
| LP | LP | Provides information about the vehicle's LP |
| Country | LP attribute indicating country of vehicle registration | |
| Property 1 | LP attribute showing the registration region number | |
| Property 2 | LP attribute showing the LP feature type | |
| Frame | LP crop. LP crop is used for attribute recognition |
A vehicle can have multiple LP crops. If a new result is «similar» to an already found number (character match above the threshold), it replaces the existing one as a more accurate result. If «not similar» (different character sequence/match below the heuristic threshold), the system considers it a different number, records it as a separate line, and shows it as a separate crop. Therefore, a vehicle may have multiple LP entries/crops.
The numbers next to each attribute represent the recognition quality of that attribute. The rating is determined within the range from 0 to 1; the closer the value is to 1, the more accurate the recognition result is.
Clicking on any frame opens extended frame information. For Track Frames, a pop-up window appears (Figure 14), containing:
- Frame ID – frame identifier
- Time – event detection date and time
- Type – frame type
- Recognition Zones – checkboxes allow showing the recognition zones on the frame
- Frame with the object - the detected object is highlighted with a BBox
Event Filter#
CARS_Analytics UI allows users to filter events to limit the display of the event list on the screen.
The filters are located on the right side of the page and are open by default. You can hide the filters by clicking on 
.
With the filters, the user can quickly find an event. By default, filters are off, and all new events are automatically loaded into the table.
Event filtering can be done by one or more parameters (Figure 15).
All filters except for Date, Event ID, Number of Axles, LP Symbols, Image Search, and Auto-fill filters support multiple selections. To cancel the filter selection, click on 
or the «Reset» button.
The filters for Scenario, Event Type, Cameras, Lists, Emergency Services, Special Transport, Public Transport, Brands, Models, Vehicle Types, Colors, and Vehicle Registration Country are presented as dropdown lists. These filters support keyboard input for searching list items. The number of items in the list will change dynamically as characters are entered.
For all filters except Date, Scenario, Event Type, Camera, Event ID, and Lists, precision settings are available. To adjust this value, click on the «Score Threshold» 
button to the right of the filter field.
When clicking on «Score Threshold»
0.000to1.000. If values below 0 or above 1 are entered, an error will appear. Also, if more than three decimal places are entered, a warning will appear showing the two closest acceptable values.
By default, all filters, except Date, Number of Axles, and LP Symbols, have an icon 
next to the field, which means the filter will be applied to the selected value. When clicked, the icon will change to 
, and the filter will be applied excluding the selected value (i.e., for all values except the one specified in the field).
The description of each filter is provided in Table 5.
Table 5. Filter Field Descriptions
| Filter | Description | Values |
|---|---|---|
| General | ||
| Date from | Time range for event registrations. Set using the built-in calendar. Enter the start date and time of the range | DD.MM.YYYY HH:MM |
| Date to | Time range for event registrations. Set using the built-in calendar. Enter the end date and time of the range | DD.MM.YYYY HH:MM |
| Scenario | Dropdown list containing event scenarios | Scenarios created by the administrator (see section Scenarios) |
| Event Type | Dropdown list containing event types | Fire – fire detection; |
| Pedestrian – pedestrian detection; | ||
| Transport – vehicle detection. | ||
| Cameras | Dropdown list containing cameras from the Cameras section | Cameras added by the administrator. |
| Event ID | Field to enter event ID | ID of any created event (e.g. e96af9df-32d5-40a6-b5cf-32e9c04632f8) |
| Lists | Dropdown list containing lists from the «Lists» section | Lists added by administrator (see section Lists) |
| Vehicle | ||
| Image Search | Search for vehicles or LPs by a given image. The search process is described in section Events. | Vehicle image |
| Auto-fill Filters by Image | Upload a vehicle image for auto-filling filters and searching by those filters. The auto-fill process is described in section Events | Vehicle image |
| Emergency services | Dropdown list with emergency services types | - Emergency – emergency service vehicles; |
| - EMERCOM – rescue service vehicles; | ||
| - Not Defined – emergency service not defined; | ||
| - Non-emergency – vehicles not belonging to emergency services (civilian, military, etc.); | ||
| - Fire – fire service vehicles; | ||
| - Police – police and traffic police vehicles; | ||
| - Other Services – vehicles of other emergency services; | ||
| - Ambulance – ambulance vehicles | ||
| Special transport | Dropdown list with special transport types | - forklift, |
| - paver, | ||
| - bulldozer, | ||
| - grader, | ||
| - roller, | ||
| - truck crane; | ||
| - concrete mixer, | ||
| - tractor, | ||
| - excavator, | ||
| - sweeper, | ||
| - garbage truck, | ||
| - sprinkler truck, | ||
| - dump truck, | ||
| - truck, | ||
| - other special, | ||
| - other non-special | ||
| Public transport | Dropdown list with public transport types | - carsharing, |
| - public transport, | ||
| - not defined; | ||
| - other, | ||
| - taxi | ||
| Number of axles | Field to enter the number of axles of the vehicle | 0…10 |
| Brands | Dropdown list with vehicle brands. Around 100 makes available | Brand available in the dropdown list |
| Models | Dropdown list with vehicle models. Around 800 models available | Models available in the dropdown list |
| Vehicle Types | Dropdown list with vehicle types | - А_light, |
| - А_heavy, | ||
| - В_light, | ||
| - B_heavy, | ||
| - С_light, | ||
| - С_heavy, | ||
| - D_light, | ||
| - D_heavy, | ||
| - D_long, | ||
| - E_light, | ||
| - E_heavy, | ||
| - P_light, | ||
| - P_heavy, | ||
| - Other, | ||
| - Undefined | ||
| Detailed description of each vehicle type can be found in the «LUNA CARS_API. Administration manual» | ||
| Colors | Dropdown list with vehicle colors | - beige, |
| - white, | ||
| - light_blue, | ||
| - yellow, | ||
| - green, | ||
| - brown, | ||
| - red, | ||
| - orange, | ||
| - purple or pink, | ||
| - grey or silver, | ||
| - blue, | ||
| - black, | ||
| - not defined | ||
| LP Symbols | ||
| LP Symbols | Field to enter LP symbols. Unknown symbols should be replaced with *. Examples of LP: A001AA777, A****A777, A000177, etc. When entering a different number of symbols, an error appears informing about the wrong number of symbols and this filter will not be used for filtering |
Numbers 0…9, Capital letters A…Z |
| LP Country | Dropdown list containing countries of vehicle registration | Countries available in the dropdown list |
The user needs to set one filter or a combination of filters and click the «Apply» button for the settings to be applied.
The list of supported vehicle brands and models depends on the selected version of the classifier for recognition (for more information, see «LUNA CARS_API. Administration manual»). When selecting the latest version of the classifier, over 160 models and 1700 brands are supported in the system.
Vehicle brand and model recognition in the system is done by the vehicle body appearance.
Vehicles of different brands and models, but with similar body appearances, are categorized under the most common brand and model in the region. For example, ZAZ Sens, ZAZ Lanos, ZAZ Chance, Daewoo Sens, Daewoo Lanos, Chevrolet Lanos, ZAZ Lanos Furgon have similar body appearances, and LUNA CARS identifies these models as Chevrolet Lanos. When the filters are set to «Daewoo» and «Sens», the filtering results will include any vehicle from the above-mentioned brands and models, but «Chevrolet» and «Lanos» will be displayed in the «Brand» and «Model» fields.
A complete list of related makes and models is provided in the «LUNA CARS_API. Administration Manual».
Clicking the icon 
to the left of the «Filters» heading opens the report generation settings window. The report will include all events matching the applied filters with the settings taken into account. Report creation happens as part of a background task. Working with tasks and applying export settings is described in section Tasks.
To reset the entered values, click the «Reset» button or 
next to the filters icon to the right of the «Filters» header.
Image Search#
CARS_Analytics UI supports searching for vehicles (Figure 16).
After clicking on 
a file upload window will appear (Figure 17).
To add an image, click on the text «Click or drag a file» and choose a file from the file explorer, or drag the file into the file upload window.
The result of the image search is a list of all events whose vehicle descriptors match the descriptor of the uploaded vehicle image with the specified accuracy. The uploaded vehicle image must meet the requirements specified in Table 6.
Table 6. Requirements for uploaded images
| Parameter | Requirements |
|---|---|
| Resolution | from 100x100 px to 1920х1080 px |
| Maximum file size | 3 Mb |
| Image color | Color or black-and-white image |
| Image Format | JPG or PNG |
| Frame composition | - Vehicle and LP must be fully visible (not blocked by other objects); |
| - Image should not contain more than one vehicle |
To reset the image, click 
in the top left corner of the image.
After adding the image, the system will prompt you to set the accuracy value for recognition. Accuracy can be deleted by clicking on the «Remove accuracy» link (Figure 18).
After clicking the «Apply» button, the image search function icon will turn blue, and the «Apply» button will change to «Create Task» (Figure 19).
When clicking the «Create task» button, a background task is created for image search.
The search result is displayed in the Section Tasks.
Autocomplete filters by vehicle image#
CARS_Analytics UI supports autocomplete filters by vehicle image (Figure 20).
After clicking on 
, a file upload window will appear (Figure 21).
To add an image, click on the text «Click or drag a file» and choose a file from the file explorer, or drag the file into the file upload window. The uploaded vehicle image must meet the requirements specified in Table 6.
The result of processing the image is the recognized vehicle and LP attributes, which can be applied to search for events. Recognizable attributes for auto-filling are:
- Brand;
- Model;
- Country;
- LP;
- Public transport;
- Special transport;
- Emergency services;
- Type;
- Color
Each attribute can be selected/unselected for use in filters by clicking the checkbox next to the attribute (Figure 22).
To reset the image, click on in the top left corner of the image.
After clicking the «Apply» button, the filters will be automatically filled in and applied according to the values of the recognized attributes (Figure 23).
