In order to improve the ability of accurate detection of pesticide residues on tomato surface, a method of nondestructive detection of pesticide residues on tomato surface based on visual image recognition was proposed. The image of pesticide residues on tomato surface was collected by laser imaging, and the spectral feature of pesticide residue was analyzed to extract the edge profile of pesticide residue area on tomato surface. Based on the feature extraction results, the region of pesticide residue on tomato surface was reconstructed by visual image reconstruction, and the partition matching technique was used to segment the region of pesticide residue on tomato surface. The detection and recognition of pesticide residues on tomato surface are realized by using adaptive block feature matching method. The simulation results show that the method has good nondestructive effect on pesticide residue on tomato surface and high information saturation of output image, which improves the ability of accurate detection of pesticide residue on tomato surface. It has good application value in tomato pest control and pesticide removal.