When the titanium alloy compressor blade of the aircraft engine works, due to the long period of highstrength service and foreign body damage, the blade will deform, dent, wear, crack and even break. Laser cladding technology has become one of the important methods of blade repair due to its small heat affected zone, good deposition performance, high forming accuracy and high degree of automation. The geometric characteristics of the melt pool are the key factors affecting the quality of the cladding, so this paper proposes an identification measurement algorithm based on image processing for the real-time monitoring of the melt pool. First of all, the ROI region is extracted by the image mask, and then the gamma transform is performed on ROI region, the threshold bination is valued to realize the segmentation of the melt pool area, the contour area feature is calculated for denoising, and finally the AABB enveloping box is used to extract the geometric characteristics of the molten pool, which realizes the real-time monitoring of the length and width of the molten pool during the cladding process. Finally, through the multi-parameter orthogonal experiment, the average recognition error of the verification algorithm is 0.24 mm.