A combined computer vision and image processing method for surface coverage measurement of shot peen forming

Abstract

Coverage is an essential monitored parameter during the shot peen forming process. However, the conventional manual visual inspection methods suffer from low efficiency and poor accuracy. In this paper, a combined computer vision and image processing method is applied to realize the automatic measurement of shot peening surface coverage, and the critical issues involved, such as shot peening image pre-processing, indentation light spots recognition, shot-peened area segmentation, etc., are systematically investigated. The results indicate that the influence of machining streaks on the subsequent image processing is effectively reduced by adjusting the illumination conditions and adopting the median filtering algorithm. The number and location of indentations in the shot peening surface image are easily obtained by treating the central light spot as an identifier for the indentation. The features of the shot-peened area can be extracted effectively by segmenting the image reconstructed based on the light spots. The effectiveness of the coverage measurement method is verified by shot peen forming experiments, and the accuracy of the measurement results can be over 97 %. The influence of coverage on shot peening deformation is further examined using the results of algorithm processing, which is reflected in that the radius of curvature decreases with the coverage.