Computer vision technology and image processing technology are applied in the field of agriculture gradually. How to diagnose crop diseases quickly and effectively has become a research hotspot. In this paper, we combine edge detection and fuzzy clustering algorithm to get the new algorithm through the experiment of more than 1500 pictures. The different kinds of diseases and insect pests of the 5 different crop leaves are used as the research object. Through the gray processing of the images, the removal of the unrelated background, the image segmentation, and the filling of the pixels of the crop disease and insect pests, the final calculation is final. The degree of crop leaf diseases and insect pests is calculated. From the experimental data, the degree of crop damage can be accurately reflected, and the degree of crop leaf disease and insect damage can be calculated, and the automatic batch operation of image segmentation can be realized.