The invention provides a skin pigment lesion treatment effect image processing method based on feature segmentation, and the method is characterized in that skin pigment lesion color images in the same region before and after treatment are processed through the steps S1 to S5: S1, converting the color images into three-channel RGB gray images; S2, coarsely extracting a focus area; S3, carrying out a repairing and closing operation; s4, calculating to obtain an area with uneven illumination of the picture; s5, low-pass filtering processing is carried out, and an actual focus area distribution map is obtained; and S6, establishing a treatment effect image evaluation model. According to the method, the focus area is segmented, the ROI area is extracted, the background light intensity of the segmented area of the image is eliminated, and the problem of background light influence in the shooting process is solved. The practical work of clinical image analysis is guided to a great extent, and a facial pigment lesion recording, analyzing and evaluating technology with high practicability is provided.