现代计算机：上半月版
現代計算機：上半月版
현대계산궤：상반월판
Modern Computer
2012年
2期
31-34,44
,共5页
灰度形态学%印刷品缺陷检测%人眼视觉特性
灰度形態學%印刷品缺陷檢測%人眼視覺特性
회도형태학%인쇄품결함검측%인안시각특성
Gray-Scale Morphology%Print Defect Detection%Human Visual Characteristics
提出一种基于人眼视觉的印刷品缺陷检测方法，方法根据印刷品缺陷检测的特点，采用减影算法检测印刷品图像的缺陷。提出动态灰度阈值的减影改进算法，此方法符合人眼对于不同灰度差别，具有不同的鉴别能力的特点。结合形态学预处理建模算法，在获得准确缺陷的情况下，能基本消除轮廓伪影，得到真实的印刷缺陷。
提齣一種基于人眼視覺的印刷品缺陷檢測方法，方法根據印刷品缺陷檢測的特點，採用減影算法檢測印刷品圖像的缺陷。提齣動態灰度閾值的減影改進算法，此方法符閤人眼對于不同灰度差彆，具有不同的鑒彆能力的特點。結閤形態學預處理建模算法，在穫得準確缺陷的情況下，能基本消除輪廓偽影，得到真實的印刷缺陷。
제출일충기우인안시각적인쇄품결함검측방법，방법근거인쇄품결함검측적특점，채용감영산법검측인쇄품도상적결함。제출동태회도역치적감영개진산법，차방법부합인안대우불동회도차별，구유불동적감별능력적특점。결합형태학예처리건모산법，재획득준학결함적정황하，능기본소제륜곽위영，득도진실적인쇄결함。
Presents a print defect detection method based on human visual. This method uses the subtrac- tion algorithm to detect the defects of the printed image according to the characteristics of the print defect detection. Proposes the subtraction improved algorithm of dynamic gray-scale threshold, and this approach is consistent with the human eye, has the ability to identify the characteristics of different gray. Based on the morphology pretreatment modeling algorithm, and in the case of access to accurate defect, can basically eliminate the contour artifacts, and get the real printing defects.
