Image Stitching and Quality Evaluation Algorithm for Large Size Parts

Abstract

Aiming at the problem of small field of view and low precision of online vision measurement for large size parts, an algorithm for image stitching and quality evaluation of large size parts is proposed.Through the establishment of imaging normalized model, calculating the transformation matrix between images, and then the camera is calibrated several times to obtain the best calibration matrix, which can realize image stitching.Finally, according to the indexes that are the SSIM value and the DoEM value of the mosaic image to evaluate the quality of the mosaic image of large size parts.Experimental results show that the proposed method can not only reduce stitching time, improve accuracy, but also meet the subjective visual perception of human, which meets the requirements of on-line inspection of large size parts.