Finding corners in an image - Harris and FAST
A corner can be thought as an intersection of two edges. The mathematical definition of the corners in an image is different, but reflects the same idea; the corner is a point with the following property: moving this point in any direction leads to changes in the small neighborhood of the point. For example, if we take a point on the homogeneous area of an image, moving such a point doesn't change anything in the local window nearby. A point on the edge doesn't belong to a plain region, and once again has directions, movements which don't influence a point's local area: these are movements along the edge. Only corners are movement-sensitive for all directions, and as a consequence, they are good candidates for objects to track or compare. In this recipe, we'll learn how to find corners on an image using two methods from OpenCV.
Getting ready
Before you proceed with this recipe, you need to install the OpenCV version 3.0 (or greater) Python API...