Silk-screen Quality Surveillance System Base On Image Processing Technology

Generally, Silk-Screen is the Glyphs which are printed on industrial products, and an important part of industrial products; Silk-Screen quality is also a notable factor that affects the quality of products.The traditional quality control method rely on the naked eye, which shortcomings of are high cost, low efficiency, poor reliability. With the fast development of image processing and printing features, this thesis presents and implements a set of Silk-Screen quality monitoring system.The main contents of this paper are shown as following:1. Build a software and hardware framework of a video capture and display system, which is the external part of monitoring system, using PC platform and the Video for Windows (VFW) function library provided by Microsoft. Implement the real-time video image capture, image format conversion, and triumphantly get the gray image which used in the core processing module.2. On the basis of study and compared the relevant algorithms, implement several image preprocessing algorithms such as a fast median filter, two-dimensional Otsu algorithm, thinning and deburring algorithm.3. Improve and implement SUSAN corner detection algorithm. A robust method of image skew correction based on feature points matching is present, which is used to skew and match the image.4. Present a distinguish algorithm base on Quasi-Euler number feature and a bidirectional relaxation matching distinguish algorithm base on Endpoint feature, and then implement these two algorithms.5. After discussing the traditional template matching algorithm and mathematical morphology, make successful application of mathematical morphology in template matching algorithm.6. Implement a complete prototype software system of Silk-Screen Quality Surveillance System.Very promising experimental and test results have confirmed that this Silk-Screen quality monitoring system has good performance.