I'm creating a program in Python, which often ends up requiring more CPU and I've researched how to use the GPU for processing. The OpenCL bases I found had C ++-only examples, and still not very explanatory. In those days I discovered that from OpenCV 3.2 there are some OpenCL functions that can be used directly by cv2.
I would like to know if this is really possible, and if so, how is the syntax to apply this throughout the code, or in a function?