摘要
提出用于缺陷检测的快速归一化互相关计算,用和表计算图像均值、图像方差和图像间的相关,使计算对模板窗口的尺寸保持不变,计算复杂度明显降低.
This paper presents a method of fast normalization and cross correlation which can detect the defection more speedily. The average value of table of image, image variance and correlation are used and the computation is not varied in the size of modular window, the computation complexity is decreased obviously.
出处
《广州航海高等专科学校学报》
2006年第2期29-31,40,共4页
Journal of Guangzhou Maritime College
基金
国家自然科学基金资助项目(50475044)
教育部科技研究重点资助项目(2004106)
广东省科技攻关资助项目(2005A10403004)
广东省自然科学基金资助项目(04300155)
关键词
归一化互相关
缺陷检测
机器视觉
和表
normalization and cross correlation
defection detection
machine vision
construction of urn-tables
