How would the FAST corner detection algorithm be implemented in C#? Here is what I understand about it. It uses the ID3 algorithm to generate candidate pixels to be analysed for corner detection, then places raster circles of a 3 pixel radius around these candidate spots. The circle production is fine, but I cannot yet understand the ID3 algorithm well enough to turn it into code that will give me candidate pixels. If anyone can offer a clear explanation of the ID3 algorithm, as to how it can select candidate pixels out of a picture, that would be awesome. Thx