Here we estimate an optical flow from a sequence of images, in other words, we will study motion estimation, that is to say, we will try to detect objects in motion (eg cars) in order to follow them, to measure their speed, etc. For this, we will implement the method of Horn and Shunck which relies on a differential approach (works for low displacements). This method of Horn and Shunck is based on the assumption that gray levels are preserved from one image to another
