Research Of QP Segmentation And Match Method For Print Image Detection

In this paper, the status of print testing were analyzed, along with the changing times and the development of technology, the spread of the printing industry as a cultural industry has been rapid development, more and more high-quality, diverse prints occupied the market. Printing has become an important part of human survival and development, people print quality and printing efficiency also have higher requirements, but existing printing technology itself still does not prevent the emergence of a variety of printing defects. How to further improve the production quality control situation, reduce production costs, increase corporate profits has become a modern printing enterprise urgent need to resolve the issue.This paper studies the status and application of the full screen print detection system, based on previous studies, focusing on the printed image QP segmentation matching method for defect detection technology. Actually detected need for a comprehensive analysis of the characteristics common printing defects, combining the principles of digital image processing is proposed based on defects in shape, size, ink thickness and color change together determine the type of defect, and described the online printing defect detection basic processes. The system can successfully detect the printing defect, depending on the image resolution and the desired detection accuracy, the type of partitioning the design image, thereby detecting and analyzing of the color information, the display characteristics of the corresponding defect, the defect types, the defect area, the relative position and timely warning eliminate waste, the appropriate feedback to the system, guiding producers to adjust the printing operation.In this paper, a polynomial regression algorithm to capture the pixel color values are converted from RGB space to the Lab space, the use of Lab color difference formulas to complete the look handicap. Finding the color characteristics of regional connectivity part, using Chebyshev distance method and pixel connected component labeling method for fastly marked defect regional connectivity, get the shape, location and size of each region based on the characteristic parameters connectivity. Since most printing defects associated with the ink thickness, this paper established by△L contact with the ink thickness, combined with the shape of the defect, a comprehensive analysis of the relative position of other features, enabling the extraction of printing defects, classified and feedback.In this paper, Visual C++to build the print line detection system, combined with OpenCV and digital image processing principle, the above method and process programming, testing and finally get print quality feedback software systems. Experiments show that the system can successfully detect print defects and the determine success of a variety of printing defects and defects in the recording area, location, print scrap, feedback defect cause of the failure to the computer interface to guide the production and defect information stored in the database.