Object_Detection_Tracking icon indicating copy to clipboard operation
Object_Detection_Tracking copied to clipboard

Is the code for real time? about the multi-camera tracking

Open alexHxun opened this issue 3 years ago • 1 comments

Thank you for your awesome work! I want to affirm if the tmot is real time? I noticed you used a for loop when dealing with multiple cameras in multi-video-reid.py: for time_slot in tqdm(camera_data["sync_groups"]): in line 448. Does you put all the frames of one video in the queue first and process track in per frame , then you handle all videos one by one?

alexHxun avatar May 24 '22 09:05 alexHxun

tmot algorithm itself is very lightweight and the object detection and tracking is very fast. The multi-threaded code in here reads the frames using CPU meanwhile running object detection on GPU (then tracking on CPU).

But for the multi-camera case, I have not tested them extensively and currently the code should be incomplete.

JunweiLiang avatar May 24 '22 13:05 JunweiLiang