GoodVision Video Insights offers two basic options for video processing to choose from, depending on the video source and level of detail required - Traffic camera processing and Drone camera processing. Drone camera has also two processing possibilities - High drone and Low drone. You can use any of your own favourite cameras regardless of the manufacturer. Please note in order to analyse extracted data properly, your cameras must have the static viewpoint.
1. Traffic Camera processing
A processing option designed for videos from fixed cameras on roads, junctions, pathways, etc. placed from 5 up to 30m high above the ground. Please note that tracking of small objects like pedestrians, cyclists and motorcyclists is reliable with lower camera placement. Videos from traffic cameras are processed in just a couple of hours from the upload. Data for all object classes are extracted from the video including pedestrians and OGV1, OGV2 recognition. For detailed specification and guidelines for traffic cameras please check Fixed Camera Processing Guidelines.
Picture 1: Traffic camera scene in Video Insights
2. Drone Camera - High drone, Low drone processing options
A processing option designed for recordings from high altitudes up to 250 metres above the ground. GoodVision Video Insights provides 2 drone processing options depending on your project needs. The processing option is chosen when creating the appropriate Data Source.
Picture 2: choosing device type by drone processing
High drone: UHD drone cameras surveying from 30 to 250 metres above the ground. This option is ideal for covering a big scene to gain a broad overview of the traffic in the area. Please note that in the high drone variant, pedestrian class is not available in the classification scheme.
Low drone: Drone cameras flying in lower heights up to 30 metres above the ground. This option is ideal for mode detailed surveys where traffic volumes and behaviour of pedestrians, cyclists or motorcyclists are desired outputs.
Drone processing can in special cases serve as an option also for UHD fixed or temporary cameras in big heights with a steep view. For detailed specification please navigate to Drone Camera Processing Guidelines.
3. Processing outputs
Processing results from will be available to you in several hours from upload, or on the next business day in complex cases (e.g some high drone scenes). In an ideal case the trajectories from a recording with a stable drone view are compact through the whole scene.
Picture 3: High drone camera scene in Video Insights
For drone videos that are not completely stable throughout the recording (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.
For such cases, we implemented a mask overlaying the camera screenshot in the Video Insights application which lets you know where these zones occurred. It is not recommended to place filters to measure traffic volumes and travel times on the grey zones (it’s possible but the outputs cannot be guaranteed). The mask is a guidance where you should be careful with the results and rather concentrate on the bright parts with complete data for your analysis.
Picture 4: Example of a dark drone mask overlaying edges of the scene
Pricing of video processing in GoodVision Video Insights
For the processing of traffic videos in the Video Insights platform, you need Credits. One Credit equals the extraction of traffic data from up to 1 hour long videos from traffic cameras or up to 15 minutes in the case of drone cameras.
For more information about Credit purchase and utilization, please visit the article about purchasing and using the Credits.
Check also our Platform plans review with all information concerning subscription of the Video Insights platform and Credit price.