| Along with the development of society and science technology, people have more requirements of printing quality,and Automotive Technology have been applied in printing.All these made it hard to test printing quality.Print quality testing is not only the basis of digital printing quality standardization, and it is essential to control follow links. The traditional methods can not satisfy the need of modern printing. Online detection method based on machine vision is gradually replacing the traditional printing quality detection methods.This paper mainly studies image mode, digital image acquisition, image pattern recognition and image registration theories,and all these are based on digital image processing theory.In this paper,an on-line detection system has been put forward.The system is based on machine vision and DSP technology.It takes DM642 chip as the core chip, using high-speed linear CCD image acquisition module and FPGA control modules to realize image collection and processing.The system adopt RF-5 structure and DSP/BIOS real-time operating system to optimize the system,which based on CCS integrated development environment.The multi-channel image of dynamic testing technology has been applied to realize real-time detection function.Section distribution in this paper is as follows:The backgrond has been introduced in the first chapter,It also involves developing situation of print quality testing technology,in the last part of this chapter we make a detailed introduction on the the main research contents of this paper; The second chapter proposes a plan of the online detection system,which are based on demand analysis;In the third chapter,we mainly study hardware system of online detection system and introduce the key part of the system in detail; The software design was introduced in the fourth chapter, and we make the simulation of image matching in the tail of the chapter;in the fifth chapter, we make experiments on the system to proof that it is effective,and the result also indicate that the system is effective rapid and stable. |