The imcorner function is an alternative to FAST's fastcorners. It is also used on corners from an image and provides a different set of methods, such as the following ones:
- harris
- shi-tomasi
- kitchen-rosenfield
Each method performs in its own unique way, and therefore feature representation will differ. To better understand the differences, we will plot them all on a single image. Let's start by loading an image of a chessboard and converting it to a float representation, as shown in the following code:
using Images, ImageFeatures, ImageMorphology, ImageViewimg = restrict(load("sample-images/board-157165_640.png"))img_f = Float16.(Gray.(img))
Next, we will be running four different methods and ...
