Application of Improved Canny Algorithm in Image of Surgical Instruments

Abstract

In order to improve the inventory efficiency of surgical instruments, a vision system needs to be installed. The edge detection of the image of surgical instruments is an important step to inventory surgical instruments. However, due to the wide variety of surgical instruments and the different sizes, it is necessary to improve the existing edge detection algorithm to satisfy the actual needs. Canny algorithm has been successfully applied in many computer vision systems, but it needs to be set manually when selecting high-low thresholds, it has poor adaptability and relies on human experience, so it cannot be directly applied to edge detection of surgical instruments. In this paper, based on the study of the traditional Canny algorithm, an improved Canny operator edge detection algorithm is proposed which shows the ability to autonomously select the threshold. Experimental results show that the improved Canny algorithm can effectively retain the true edge information, make the edges more continuous, and effectively achieve the edge detection of surgical instruments compared with the traditional Canny algorithm and the multi-scale morphology algorithm.