Research and Application of Image Edge Detection Algorithm Based on FOA

Abstract

Traditional differential-based edge detection algorithms have shortcomings such as inaccurate edge point positioning and corner point missing detection. This paper presents an application of FOA (Fruit Fly Optimization Algorithm) for establishing an image edge tracking model. Firstly, the prior knowledge of edge points is obtained by Canny edge detection operator. Then the corner points are extracted by Hilbert transform, and the edge tracking is accomplished by the coordination of the two mechanisms of randomness and positive feedback; Finally, a single pixel edge is obtained by using the forward movement mechanism and the loop termination condition. Compared with the original Canny edge detection algorithm, the precision and accuracy of image detection are better