What are the advantages and problems(disadvantages) of Multi Target Tracking using a single camera sensor?
Dear Ramin,
Thanks for your question. I have attached few files for your look at it. I think these files will answer your concern in a detailed manner. Feel free to ask if you have more doubts.
Regards,
Dr. Kar
Hi, I suggest : http://motchallenge.net/
In this website you can find your answers and connect to the experts to improve your knowledge.
Be successful
It depends upon your area of application, we have successfully tracked multiple skin regions (hands and faces) for multi-gesture recognition using a Logitech Webcam, However, feed from a single camera may not suffice in application areas like surveillance or analysis of sports videos.
I agree with @M Raza Ali , any application with the goal of taking accurate positions of spatial objects will need more than one camera. The advantage is the tracking complexity, with one camera, you have to apply some algorithms for detecting and computing the pose, but if you use more than one camera, you will have to do the match between the corresponding images.
Look for the book: Multiple view geometry of Hartley and Zisserman
Hello. With multiple sensors you have multiple fields of view assuming the sensors are reasonably well separated relative to the range of the opjects. With a single sensor you have a standard multi-target tracking problem, as covered in the reference below. With multiple sensors you also have to perform data association across the different sensors as well as between the objects seen by one sensor. Another reference is supplied from the radar domain for this problem. With multiple sensors, you may be able to avoid occlusion.
Article Taxonomy of multiple target tracking methods
Conference Paper Simultaneous registration and tracking for multiple radars w...
I think there have been a lot of good answers posted to this question. I'll add my opinions/experiences that will hopefully act as a summary.
Advantages of multi-camera tracking:
-Could resolve occlusion problems where a target is not visible in one camera (due to occlusion) but is visible in another camera
-Could produce a more accurate geometric understanding of the scene (3D locations and velocities, range to each target)
-Could cover a wider area (and therefore more targets) than a single camera
-Potentially more accurate tracking results through the use of multiple views for better target identification
Disadvantages of multi-camera tracking:
-Need to solve the data association problem where we determine the correspondence between targets in different cameras. This could be solved with calibrated cameras but I don't think that's easy to do.
-Need to determine how data is going to be sent between cameras. Will the data be collected at a centralized node and processed there? Processed at each camera individually and results transmitted to centralized node? Distributed fusion to remove the need for a centralized node? Transmitting video data can require lots of bandwidth.
-A multi-camera system will probably be more expensive than a single-camera system because of the extra sensors and the supporting infrastructure.
- Badges
- Science topic
- Similar topics
- Chemistry
- Biochemistry
- Biochemical Techniques
- Blotting Techniques
- Immunoblotting
- Dyes