Research on Online Detection Algorithm for Printed Matter Based on Image Processing

摘要

Online detection of defects in printed matter has practical significance. In this paper, compositionand operating principle of the detection system are introduced. To solve the problem of slow computationspeed and low accuracy in image matching, a new approach using adaptive genetic correlation matching algorithmis proposed. Correlation matching method is widely used in image matching, to save computation timeadaptive genetic algorithm (AGA) is imported. The advantage of AGA is that the probabilities of crossoverand mutation are varied depending on the fitness values, thus improving the performance in searching speedand global convergence. Based on Visual C# software development platform, template image constructionmodule, image registration module and image analysis module are designed finally. Experimental resultsshow that the proposed image registration algorithm has high accuracy in matching and time-consuming is inless than 100ms, which reaches the design requirements of online detection system.