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

Tracked object shows live_points as True even though there is not detection in the current frame at that coordinate #328

Open
devvaibhav455 opened this issue Jan 24, 2025 · 0 comments
Labels
help wanted Extra attention is needed

Comments

@devvaibhav455
Copy link

Describe the situation you are working on
I am running tracker on detections obtained by an object detector.

For this particular frame, I have 10 boxes whose centroids are passed to the norfair tracker which gives me 13 tracks. Now, I expect that some of these tracks won't be associated with the detection boxes as the tracks are more in number than the detections. I think that live_points tells us about that property. If it is True, it means that the track is matched with a detection, otherwise not. But in this particular case, live_points is True, current_min_distance is not None and its plotted in the sky. I tried checking the object detector model's raw output too along with the visualization but there isn't a box at the place where track number 91 is suggesting.

Box center is calculated as below from the filtered_detections passed to norfair as below.

for i in range(len(filtered_bboxes)):
            box = filtered_bboxes[i]
            conf = filtered_confs[i]
            cls = filtered_classes[i]

            points = np.array([(box[0]+box[2])/2, (box[1]+box[3])/2]) #Track the center
            
            # Assign scores (you can use actual detection confidence if available)
            scores = np.array([conf])  # Example: Default confidence for each corner
            data = np.array([box[0], box[1], box[2], box[3], conf, cls])
            
            # Create a Norfair Detection object
            detection = Detection(points=points, scores=scores, data=data)
            norfair_detections.append(detection)

coord_transformation = self.motion_estimator.update(self.new_image)
output_tracks = self.tracker.update(detections=norfair_detections, coord_transformations=coord_transformation)

Describe what is it that you need help with
The tracked object of interest is with id: 91
Below is the stack trace output from the VSCode debugger.

Image

Image

Image

Image

Image

Image

Image

Additional context
Add any other context about the problem here.

@devvaibhav455 devvaibhav455 added the help wanted Extra attention is needed label Jan 24, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
help wanted Extra attention is needed
Projects
None yet
Development

No branches or pull requests

1 participant