基于机器视觉的印刷品质量在线检测系统研究-机械制造及其自动化专业论文.docxVIP

山东理下大学硕七学位论文 山东理下大学硕七学位论文 Abstract Abstract By stucling on the advantages ofmachine vision applied to industry detection and wavelet transform applied t0 inlage processing,thc real-time printing quality detection system is developed aimed at detecting the printing quality． Firstly,this thesis studies the image capture ofprinting quality detection system which is focused on liner CCD,and illlloduces the principles and恤applications ofl斌1ware used to capture inlage,3．In order to avoid the influence of e玎_or factors such as noise and geometry distortion,a method of image de-noising based on wavelet threshold shlinkage is proposed after analrzing various de-noising methods．仳method c锄de-noise efficiently and豫seⅣe the edge information as much Eis possible．Atter de-noising,the image shoud be∞gis删for the image matching． Secondly,this thesis compares several Iraditional matel崦algorithms and analyzes their respl碰ve merits and demerits．Considered both accuracy and efficiency for matching,ft．new matehiag method is presentⅨl,whiela combines the double-template matching and wavelet muld-resoludon layer matching．憾lllelhod c趾eflfieiently avoid mismatching and greatly reduce the calculation,so the ett]eieney and the accuracy of printing quality detection system can be satisfied． Finally,the thesis puts forward fill overprinting detection algorithm based oil image information．m image eaptured by liner CCD should be pre-proeessed,eolor-seperated and image-matched,then the overptil衄error锄be computed by the image-matching data．As the experiment shown,the overplJl衄system has the possibility and superiority of the real-time pI缸血lg quality det枷on system,so it c雒satisfy the anticipated goals．It supports lnore theories and experiment data to further research ofthe real-time printing quality dd[ecd∞ system． Key Words：machine vision，in3age capture，image de-noising，wavelet Ir珊lsform, image matching Ⅱ 山东理工大学硕十学位论文 山东理工大学硕十学位论文 第一章绪论 第一章绪论 1．1引言 近二十年来，我国印刷工业稳步高速发展。进入21世纪后，电子媒体和因特网更强 的时效性和更多的个性化内