I am trying to capture each cv.jit.faces into buffers and composite all into a matrix with accumulating alpha value by alphablend and a timer.
the problem is when it detects multiple face at the same time, its order is changing by each bound sizes. so captured bound that is in process of adding alpha value is stay still but the face switches with another one.
how can I sorting cv.jit.faces to prevent mix up.
second question is.
when I composite each buffer images into one matrix, buffer image is deleted when a new face of the slot detects.
how can I name image buffers, so I can keep previous buffer images that will be added into the composite matrix.
plus, is there a way to add (additive) each buffer images with a compositing matrix that contains previous images, without using jit.op + each buffer images?
re first question: I could never get the hang of this either, sorry I can’t help but I’m interested to know the answer to keeping faces sorted.
I think in premise you could probably recognize a face then track it and use the cv.jit.blobs.sort (I think that’s what it’s called). But I don’t get very far along before giving up.
C74 RSS Feed | © Copyright Cycling '74