基于机器视觉的印刷图像色彩在线检测系统设计  

The Design of Online Detection System for Printing Image Color Based on Machine Vision

作　　者：李文育[1] 张二虎[1] 于佳[1] 陈亚军[1] 

机构地区：[1]西安理工大学印刷包装工程学院,陕西西安710048

出　　处：《西安理工大学学报》2013年第2期159-164,共6页Journal of Xi'an University of Technology

基　　金：陕西省科技发展计划基金资助项目(2008K07-20)

摘　　要：对印刷图像色彩质量进行在线检测,可以克服人工检测速度慢、检测质量不够稳定的缺陷。针对印刷生产实际过程的需求,首先研究了检测系统的硬件平台设计并对硬件系统的参数选择进行了分析,然后对系统中的图像配准、标准模板建立及色彩空间转换等方法进行了研究。从减少检测时间的角度,提出了粗检与细检相结合的检测策略及基于边缘特征的小邻域多区域模板配准方法。实验结果表明,所设计的检测系统可以达到在线检测系统的要求。Detecting the quality of the printing image color online can overcome such defects as the low speed of manual detecting, the instability of the detecting quality. Firstly, the hardware platform design of detection system is studied according to the actual printing production process requirements, and pa- rameters of hardware system are analyzed. Then several methods have been researched including image registration, the standard template establishment and color space conversion. From the point of reducing the detection time, the detection strategies about combination of coarse detection and closeIy screening, as well as the edge feature template registration. Experimental results showed that the designed detecting system can meet the requirement of online detecting system based on the small area are suggested in this paper.

关 键 词：印刷色彩 机器视觉 图像配准 色彩空间 

分 类 号：TS807[轻工技术与工程]