Research On Print Quality Online Inspection Technology

Recently, with the development of modern technology and information standards, in terms of industrial requirement, or the people's appreciate levels, the requirements of the print quality are increasing, while the color of the image is an important message, the quality requirements for color image are more and more strict. Under this environment, the print's color quality inspection techniques were studied.In the process of print quality online inspection, the picture is acquired by the CCD camera, the collected images are RGB color mode, and its value can not to be used to judge the color quality. So in my study, I choice a uniform,device-independent color space as the objective color space to evaluation the color quality. I choice CIE1976L*a*b* to establish the conversion model from RGB color space to CIE1976L*a*b* color space.Based on the study of color space conversion methods, this article choice the color space conversion model which established a three dimensional look-up table based on tetrahedral interpolation. To the model I studied, I composed the arithmetic by C++6.0, the program achieved the conversion from RGB to CIE1976L*a*b* color space. And then the experiment verified the accuracy of the color space conversion model by the test dates. The experiments show that the look-up table established by nine levels separate has higher precision. It can be used for the color space conversion in the process of image color quality inspection.Based on the establishment of color space conversion model, the experiment design the color quality detection module based on the software C++6.0. The experiment get and display two BMP images successfully, and obtain the color date of the image. One of the two images is the standard image, the other is the image to be detected. The dates are conversed to the values of target color space, then evaluate the color quality of the image by CIE1976L*a*b* color difference formula. At last, the location and color information of the defective points are displayed.