Image segmentation of overlapping leaves based on Chan-Vese model and Sobel operator

To improve the segmentation precision of overlapping crop leaves, this paper presents an effective image segmentation method based on the Chan-Vese model and Sobel operator. The approach consists of three stages. First, a feature that identifies hues with relatively high levels of green is used to extract the region of leaves and remove the background. Second, the Chan-Vese model and improved Sobel operator are implemented to extract the leaf contours and detect the edges, respectively. Third, a target leaf with a complex background and overlapping is extracted by combining the results obtained by the Chan-Vese model and Sobel operator. To verify the effectiveness of the proposed algorithm, a segmentation experiment was performed on 30 images of cucumber leaf. The mean error rate of the proposed method is 0.0428, which is a decrease of 6.54% compared with the mean error rate of the level set method. Experimental results show that the proposed method can accurately extract the target leaf from cucumber leaf images with complex backgrounds and overlapping regions.