Application Of Image Detection In The Chromatic Printing Control System Based On Wavelet Analysis

During the intaglio printing, in order to ensure the quality of the printing product, we must fit the machine for detecting the error. Common methods of detecting error are that, such as detecting using photoelectron, detecting by worker observing the mark etc. Using the first method, the course can not be observed. Using the second method, the process is not automatic. In this article, the author delivers a detecting method based on the image analysis. This method is automatic and can be observed in the process. In the method, at first we capture the mark for the detecting using CCD digital camera, then the image that has captured is transferred to the computer. The image is displayed on the monitor and is analyzed in the computer memory. In the article, the author uses the wavelet analysis to deal with the image. The main content in the research are as follows.At first, the research of the intaglio printing principle and the status and the theory of the error detecting are discussed. Secondly, the process of some digital image is developed, and the emphases is on the image enhancement and segmentation. Thirdly, the research of the wavelet theory and the application in digital image process are discussed. Fourthly, the experiment plat of the hardware and the software of the detecting system is analyzed. And a method of detecting error is designed. Finally, the arithmetic of the enhancement based on the histogram and the segmentation that the multiscale edge detector based on the local modulus maxim of a wavelet transform are discussed.The error detection based on the wavelet analysis that is discussed in this article is the experiment and the exploring of application of digital image analysis in the intaglio printing detecting error. In the experiment flat, the system has be tested, and the precision of the detecting error is 0.065mm and satisfied the requirement of the plan. The software of the method has been tested by MATLAB and is programmed in Borland C++ in DOS OS. Research result of this paper enriches the connotation of image analysis system.