Abstract: During the robot welding, the requirement for clarity of the weld image is high to facilitate the identification of the weld, but the higher the clarity is, the higher the cost is. Therefore, an enhancement algorithm to obtain a high-definition image at low cost was proposed. Firstly, the weld image is pre-processed, and differential filtering between Gaussian filtering and bilateral filtering is used to process the original image, which can effectively reduce the impact of noise and preserve the edge details of the image. Then, the CLAHE (contrast limited AHE) algorithm is used to enhance the pre-processed image to improve the image contrast, and the low-frequency information after noise reduction is obtained by using Gaussian mask processing, and then the high-frequency information of the weld image is highlighted by linear difference between the CLAHE algorithm result and the Gaussian mask processing result. The image contrast is further improved by linear addition with the CLAHE enhanced image, and the enhanced image is finally obtained. By comparing with other algorithms, the algorithm not only improves the contrast, but also obtains clear weld edges, which proves the effectiveness of the algorithm.