1 | initial version |
You could use the blob tracking approach (samples/c/blobtracking_sample.cpp) and documentation here.
The other interest points could be another solution for your problem if you could model your object. See Features2d for the one in OpenCV (SIFT/SURF/FAST/ORB/...). Their is also some tutorials that can help you here.
After you are able to identify your object, you could also improve the tracking with the Kalman filter (see samples/cpp/kalman.cpp) for an example.
If you are tracking moving object, what ever they are, look at the background subtraction methods here and the samples in cpp/.