As you know, GoodVision Artificial Intelligence algorithms automatically extract all the physical aspects of vehicle behaviour from the video footage so that you can get all kinds of metrics and visual maps for your further traffic analysis. The only thing to do after your video is successfully processed is to - Define the conditions on the scene. Whole process is simple, you need to do it only once per each camera. After you're done, you can upload more videos to your camera and get the traffic data easily via a few clicks to your widgets' dashboard as the settings and filters are already in place.
STEP 1 - Place the virtual zones and lines
On the "Describe page" draw the virtual lines or zones to all intersection approaches or lanes you want to include to your traffic analysis. Describe scene displays the exact traffic trajectories which will help you to draw zones or lines to the right places. Make sure you draw the zone or line that capture all the trajectories on the given approach, or you will miss some and you will get lower counts!
The image below show the example of how the zones can be placed. It is one of the ways to do it.
STEP 2 - Define Traffic Movements, Events, Scenarios
The most common way how to define your traffic behavior is by defining the traffic movements. Movements allow you to filter traffic objects that are driving their path. This is necessary as the baseline for all kinds of analyses, from traffic counting to advanced analyses of travel time, saturation flows and more.
Example below filters traffic objects (trajectories) that were crossing the zone1, then zone2 and then zone 3 in that specific order. For more information, click below and read the full article describing traffic movements in detail.
For drone videos processed by high-detail drone processing which are not completely stable, e.g. slightly moved, rotated or jittering, we have developed a stabilization procedure that helps to hold the scene still so that the trajectories remain as compact as possible. However, it can easily happen that some parts of the video were not in focus all the time during the footage, as the drone would turn a bit or change its position. That influences the analysis negatively, as the number of vehicles in the "negative zones" would necessarily be lower than in reality, just because they were not in the camera view for some time.
For such cases, we implemented a mask overlaying the camera screenshot in the Video Insights application which lets you know, where these zones occurred. You will still see the objects' trajectories there and can still put lines and filters to measure traffic volumes and travel times on the grey zones but it' s not recommended because the outputs will not be complete by definition. You can consider the mask as guidance where you should be careful with the results and rather concentrate on the bright parts with complete data for your analysis.
The second option for defining specific traffic behavior conditions is using Events. Events allow you to filter traffic objects by various behaviour (e.g. by occupancy time, only vehicles that stopped in the area, and more). Events allow you to create advanced conditions to your filters and scenarios (read below).
Example below filters traffic objects that were staying in the middle of the intersection (zone5) for 5 to 35 seconds. For more information, click below and read the full article describing events in detail.
The third option for defining the traffic behavior is to create a traffic scenario that allows for combining various movements and events into complex conditions. Scenarios use simple logic operators AND, OR, THEN, and NOT. With this tool you can easily filter the traffic trajectories for the most complex traffic situations you can think of.
Example below filters traffic objects that followed one of the defined traffic movements AND were occupying the middle of the intersection for more than 15 seconds. For more information, click below and read the full article describing traffic scenarios in detail.
STEP 3: Getting the traffic data reports
Finally to analyze the traffic data defined by your movements, events or scenarios, create appropriate widgets on the Analyze page or reports using the Report Builder. See our guide on How to create widgets for the traffic volume, travel time, peak periods or various visual maps to get the most of the Video Insights platform.