Perhaps the question is not entirely correct, but still What is the difference between OpenCV ML and Tensorflow? What is better to apply?

    1 answer 1

    The question is too general and the choice depends entirely on the tasks. Libraries are not just different, they have a completely different purpose. In short:

    • OpenCV contains implementations of various machine vision algorithms. Methods of machine learning there are implemented only a couple of pieces.
    • Tensorflow is a library for working with general purpose tensors, without being tied to a specific area. But it implemented autodifferentiation, optimization methods and the ability to use the GPU to compute arbitrary complexity.

    If your application only needs to find a face in a photo or using clustering to reduce the image palette, then it’s easier to take OpenCV. There it is done in a couple of lines.

    If you are teaching the neural network of the original architecture for digital image processing, then, of course, you will have to use Tensorflow.