Image matching algorithm of defects on navel orange surface based on compressed sensing

Abstract

In this paper, an improved sub-pixel edge detection algorithm combining coarse and precise location is proposed. The algorithm fully considers the 8-neighborhood pixel information and keeps the Roberts operator’s advantages of high location accuracy and fast speed. Meanwhile, it can effectively suppress noise and obtain better detection results. In order to solve the problem of low efficiency of the Zernike moment method in threshold selection, the Otsu’s method is introduced to achieve accurate sub-pixel edge location. The experimental results show that the proposed algorithm effectively improves the detection efficiency and the detection accuracy.