Haar Cascade instability
Hello,
I trained my Cascade Classifier to identify an object. It works, but not in all frames.
Here is a frame missing one object:
And here is other frame with all the objects:
The images are almost the same. When I apply the cascade in a video, I can see the moments that the objects are not identified and it looks incomplete. Does anybody know how can I identify all the objects in all the frames, so the green rectangle will be permanent? Why it is happening, since the images are almost the same?
Any tip will be very helpful,
Thanks.
Cascades are sensitive to light changes and thus summation of pixel changes. In a video in a supermarket, where lighting pulses at a certain frequency, this can influence the different frames, as well as the auto setting functions of the camera grabbing the frames. To avoid this, filtering in time is the only way to go, which can be done as suggested below by either a Kalman filter or by an Optical Flow filter!