The cv.jit.faces tracking works great for me but not from a long distance.
I am testing to use camera matrix with higher dimension and tracking only the part with the face(s). My question:
Does somebody know a better solution/ideas to do face tracking with cv.jit.faces from a long distance? (i mean with this; a camera that is far away from the people to track, so that the faces are always very small in the matrix)
Hi, the smallest face that cv.jit.faces can recognize is 20 by 20 pixels. You need to adjust the resolution of your source image in consequence. Your only options are to use a narrow angle lens, or a high resolution source.
That is the way that i (bad) described; "camera matrix with higher dimension" and that i already use. I will test the narrow angle lens solution. Thanks
btw: thanks for your amazing cv jitter external collection, i use them with pleasure for many years now.
C74 RSS Feed | © Copyright Cycling '74