Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

missing objects for tracking objects video= 10fps #146

Open
PROGRAMMINGENGINEER-NIKI opened this issue Oct 15, 2021 · 3 comments
Open

missing objects for tracking objects video= 10fps #146

PROGRAMMINGENGINEER-NIKI opened this issue Oct 15, 2021 · 3 comments

Comments

@PROGRAMMINGENGINEER-NIKI
Copy link

PROGRAMMINGENGINEER-NIKI commented Oct 15, 2021

Thanks for the useful repository.
I combined the sort algorithm with an object detection(YOLOV3), I run the application at 30 fps but my video feed is 10fps. I noticed when the video feed is slow,10fps, the system does not work well, it misses some object for a few frames and id switching happens.
I wonder why this is happening? is it due to the Kalman filter? How Kalman filter can be affected when the video is slow 1 or 10 fps?

@cappittall
Copy link

@PROGRAMMINGENGINEER-NIKI hi, I have the same issue. Have you solved the problem? I tried to change --max_age argument which keeps alive the traking number even not detected at next frames. But seams not working as I expected.
(max-age - Maximum number of frames to keep alive a track without associated detections.)

@Ashsur
Copy link

Ashsur commented Jan 18, 2022

@PROGRAMMINGENGINEER-NIKI hi, I have the same issue. Have you solved the problem? I tried to change --max_age argument which keeps alive the traking number even not detected at next frames. But seams not working as I expected. (max-age - Maximum number of frames to keep alive a track without associated detections.)

in Line245: if (trk.time_since_update < 1) means kalmanTracker must [update] before [predict],so,if the detector output FrameRate is slower than video FrameRate, SORT will not output correct.

solution:

  1. change if (trk.time_since_update < 1) to if (trk.time_since_update < VIDEO_FRAME_RATE/DETECTOR_FRAME_RATE);
  2. chage max-age=1 to max_age=VIDEO_FRAME_RATE/DETECTOR_FRAME_RATE

@varungupta31
Copy link

varungupta31 commented Jan 27, 2022

Hey @Ashsur @cappittall @PROGRAMMINGENGINEER-NIKI
I have generated a sort output file for my custom video dataset, and want to visualize the tracker output on my custom video (not on the MOT benchmark data). Are you aware how to do this?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants