In view of complex background of lesion images and the difficulty in segmentation, an improved C—V model based on level set and weighted color information was proposed and applied into agricultural lesion image segmentation. The segmentation model of weighted color information based on level set was suitable to different diseases identification and could identify lesion disease automatically. Experimental results show that the proposed model has better property than C—V model, and has many advantages such as anti-noise and scalability properties on 4R—G image model for cucumber pink and 3R—G—B image model for apple rust disease.