Research On An Improved Edge Detection Algorithm

Digital image processing technology is widely used in real life,mainly in industry,medicine,aerospace,communications engineering,and modern agriculture.In this paper,the edge of the image is obtained by edge detection technology,and the maturity of watermelon is used to judge the edge of the image.Based on the research of image de-noising and image enhancement in digital image processing,this paper focuses on the advantages and disadvantages of traditional edge detection operators and classical Canny algorithm.Based on the classical Canny algorithm,an improved algorithm is proposed.The improvement is mainly reflected in two aspects: First,make full use of the principle of arrangement combination,through the improved median filtering noise;Secondly,the best threshold is calculated by combining the Otsu algorithm and the threshold iteration method to detect and connect the edges,and the initial threshold is defined by defining the gray value set after non-maximum suppression,and the threshold threshold is adjusted according to the segmentation results until it meets the requirements.The simulation results show that the improved algorithm proposed in this paper is used for edge detection.The results are better than classical Canny algorithm,edge detection algorithm based on bilateral filtering,edge detection algorithm based on anisotropic diffusion equation and edge detection algorithm based on robust principal component analysis.The experimental results show that the edge information detected by this algorithm is clear and complete,the pseudo edge is relatively small,and the image processing effect is better.