Since version 19, GoodVision Video Insights supports the speed estimation of traffic attendants. Read more about the video requirements in order to use this powerful feature in GoodVision Video Insights.

What are the video requirements for speed estimation?

Overal video quality requirements are derived from the standard guidelines described in Standard Camera Processing Guidelines. In addition to these, to assure the best accuracy when estimating speed, there are additional requirements for the video you should follow before uploading the video.

  1. Video scene, where the monitored vehicles move, should be flat. Scenes, where vehicles are driving through hills, up and down, or otherwise change their distance from the camera,  are not supported.
  2. Camera must be static during the recording and the position of the camera itself should not be rolled - keep the horizon line leveled. See image below for better comprehension.

Examples of supported camera positions - pan and tilt and not supported - roll

3. Beware of the barrel distortion of the camera and don't use fish-eye lenses.

4. Scene should contain visible parallel straight lines of some kind (traffic lane markings, sidewalks, fences, building edges,...) which greatly helps us to calibrate the scene.

5. Ideally, camera should be angled more towards ground (rather than perfectly horizontal) and placed rather higher above the scene

NOTE: Speed feature is not available when processing video streams.

You should also know camera attributes before the upload

Before uploading, check your camera attributes which help us to calibrate the speed estimation algorithm. Specifically:

  1. Height of camera above the scene (this can be bypassed by specific measurements in the scene, if camera height is not known)
  2. Manufacturer and model of the camera, including the focal length used if zoom is available on the camera. 
  3. Location of the camera (GPS coordinates) (optional)

You have to select if you require speed estimation when you are uploading your video

Speed estimation is one of the post-processing options which are available for both types of processing - Standard and Drone high-detail. However, the speed estimation is currently in BETA and therefore, you should contact the support via our live chat to enable this feature for you. After you have this feature unlocked, you just need to check the speed estimation checkbox before the upload. 

Example of video upload dialog with the marked speed estimation option

How will I obtain the speed report ?

Currently, the process of generating the speed report is semi-automatic. Therefore, after you upload the video with the checked speed estimation option, we will reach you via live-chat or email to learn about the attributes of your camera mentioned above.

After that, we will configure movements in the scene based on your needs and run the speed estimation post-processing on the video. As a result, you will receive the selected speed report type in the CSV or Excel format to your mailbox. Typically, this report will contain vehicle entry time to the movement, vehicle exit time from the movement, vehicle class and average speed within the areas specified. See the snippet from the report below.

Example of one type of speed report for vehicles travelling from zone 1 to zone 2

Can you estimate speed on videos which were already processed on my account in the past?

Unfortunately, no. The speed estimation is done as part of the post processing step so in case you need the speed of vehicles for your traffic analysis, please upload the video again with the checked speed option to get the results.

Did this answer your question?