Defect detection of printing images on cans based on SSIM and chromatism

In the production of digital printing equipment, it is a challenge to detect the printed surface defects of cylindrical surface like cans. This paper presents an automatic inspection program based on computer vision. We build a set of printing detection hardware system including the automatic image acquisition, high precision image registration. Then we propose a geometric defect detection based on SSIM and color defect detection algorithm based on CIEDE2000 color difference formula. The experimental results show that the system proposed by us has good effectiveness and robustness, and can realize digital detection and feedback adjustment.