Overprint Alignment Technology Research Based On Machine Vision

In a colored printer printing process, the accuracy of the overprint is a key factor affectingprint quality. The overprint is the printing process which images of different colors areprinted to a specific location. In order to obtain high quality printed image，it is necessary toensure that different colors must coincide with the monochrome image accurately. If theoverprint is not accurate, color printing can not be finished, because the overprint accuracyaffects the fineness of print and harpness of the picture, affects the quality of the product.Thiswill have a considerable impact on the product packaging process. If the overprint is accurate,the rejection rate in the production process can be greatly reduced, which can save a lot of rawmaterials. The production efficiency can be greatly improved. To detect the accuracy of theoverprint, we often use "positioning notation". The mark is printed on the production in aspecific location. By examining the degree of overprint of the mark,we can determine theoverprint error in order to adjust the printer to achieve high quality printing. At present, thedegree of automation and precision of the automatic overprint detection systems abroad is high.But the systems is expensive, and the core technology is also confidential. China also carriedout research related to the technology, but there is no full-form products. Most printingtechnology of the domestic production are imported. Therefore, overprint technology researchhas important significance for China’s printing industry.Machine vision has been widely used in many fields. It can increase the degree ofautomation, so that some industrial activities could become more flexible, reduce the externalconditions for productive activities constraints and limitations. The technology, using machinevision to collect specific flag, measuring overprint error with digital image processingtechnology, becomes a trend. In this paper, the overprint error detection system which usingmachine vision technology were studied. Firstly, the current status of the overprint wereintroduced, and machine vision technology were also overviewed. Secondly, it introduced the composition of the overprint error detection machine which using machine vision system. Theselection of camera, lens and light source for image acquisition was also carried out. Next, itintroduces a method for detecting overprint error, including pattern design for the overprint errordetection, image pre-processing, template matching algorithms, and related experimental results.Overprint error detection process involves: different shape was designed for each color as thedetection icon; different colored icons were arranged according to certain rules; importingtemplates, capturing an image with all the icons, followed by the selection of each icon as atemplate; and finally calculating the coordinates of each template, which can compute overprinterrors.In this paper, the overprint error detection techniques based on machine vision was studied,we propose a solution to overprint error detection scheme that can achieve detection of overprinterror, which has very important significance to the improvement of the printing industryoverprint error detection techniques.