A New Robust Distance Machine Leaning Approach to Edge and Corner Detection: RD-FAST

Abstract

According to [1] the RD RANSAC gives better performance of other SAC procedures and associate with the concept of SFAST [2]. SFAST method gives better performance compared to FAST [3]. In this paper, to contribute a new approach in Robust Distance (RD) SAC to use in computer vision in specially edge and corner detection a technique is defined as RD-FAST. SFAST is based on S estimator, it’s used in standard deviation is the choice of distance. In this usual standard deviation is not performing well the data is irregular. To overcome this concept we use new approach of robust distance in FAST algorithm (RD-FAST). It is performing better compared to FAST and other FAST related methods. All experimental is carried out MATLAB software.