cv.jit.faces save bounding box

Qbrick's icon

Hey Peeps,

Can somebody tell me how I could save the image in the bounding box with the cv.jit.faces example.

Preferably with a timestamp and in two differen sizes 32X32 & 128x128. I could then use those selected faces for a training model in openface or dcgan.tensorflow.

Also Is it possible to have the cv.jit.faces recognize multiple diffrent faces and save them seperately?

@jmpeltier are you still around with the your awsome cv.jit work?! :)

Rob Ramirez's icon

you can export as jpeg or png, and resize both using the jit.matrix object. explicitly set the properties as args (e.g. [ jit.matrix 4 char 32 32 ] or whatever your properties are) and it will resize when you send the matrix in. you can then send the exportimage message to export to file.

Qbrick's icon

But how can I get the images that's in the box of cv.jit.faces.draw?

Thanks!