Noise removal in industrial robot image processing is an important and challeng-ing task.Traditional image filtering algorithms have limitations in handling different types and in-tensities of noise in industrial robot images.In this paper,we propose a mask-base-d image filte-ring technique for industrial robot images.By using mask segmentation and a combination of multiple filtering algorithms,this technique effectively removes various t-ypes of noise in in-dustrial robot images and improves the accuracy and stability of image processing.Experimental results demonstrate that the proposed method achieves good perf-ormance in removing salt-and-pep-per noise and Gaussian noise,providing a viable solutio-n for industrial robot image processing.