Abstract:An improved Grab Cut image segmentation algorithm was proposed to solve the problem of inaccurate segmentation results caused by uneven lighting in the workshop during unordered sorting.In the algorithm,adaptive median filter was used to remove the image surface noise.By limiting contrast adaptive histogram equalization,the overall brightness of the image and the contrast were enhanced;Laplacian operator was further used to sharpen the brake disc area;on this basis,the complex background of the brake disc image was removed.Finally,the algorithm model was evaluated by comparing the segmentation effect,PSNR,segmentation error rate,segmentation accuracy and running time through simulation experiments.The experimental results show that compared with the original algorithm,the proposed improved Grab Cut segmentation algorithm can be used to effectively reduce the influence of uneven light on the removal of complex image background of the brake disc,and the segmentation accuracy is increased by 12.5%.The segmentation result has a small error compared with the actual brake disc model,providing reference for the removal of complex image background of the brake disc.