An Improved Adaptive Threshold Canny Edge Detection Algorithm

Abstract

It has proposed an adaptive threshold edge detection algorithm in this paper, which applies the bilateral filtering that has the advantages of edge-preserving and noise-removing firstly. Then it uses OTSU, which is based on gradient magnitude to maximize the separability of the resultant classes, to determine the low and high thresholds of the canny operator. Finally, the edge detection and connection are performed. The experimental results show that this algorithm is practical and reliable.