A corner detection algorithm based on the difference of FCC

Corner detection is important in image process and pattern recognition. In this paper a fast algorithm for corner detecting is proposed. This algorithm is based on Freeman Chain Code(FCC) and it can discard most of the possible corners by computing the sum of the difference, so by calculating less points of curvature, the corners are easily get. Meanwhile this algorithm has strong robustness for the noise in the line,the algorithm can effectively filter the noise points without denoising, it is high accuracy, low computation and easy realization in hardware.