numpy-opencv-converter | OpenCV < = > NumPy Converter using Boost : :Python | Machine Learning library

I think you get a "bytes object" if you do a read() on the result of tar.extract file(). Then you can convert that to a Numpy ndarray and use OpenCV imdecode() to unpack the TIF format from the memory buffer into an image.

This code works for me - though I seem to alternate between reading 266 byte useless files and TIF images from my tarfile - I don't know why that is but if you don't get them, you can just remove the if statement that checks for useless 266 byte files!