摘要
对织物表面出现的断经、断纬、破洞、油污等疵点进行识别并在实际中应用。将熵阈值分割应用于图像处理,通过最大熵阈值分割的迭代运算,将目标区域与背景区域分割开,即将织物疵点区域与正常区域划分出来,然后进一步对图像进行特征化处理,同时将实际生活中的乞丐装样式织物进行特征化处理,经过二者之间的特征化图像比对,确定疵点织物应用于实际生产的样式,以实现对疵点织物的有效利用。
The defects such as holes, warp-lacking, weft-lacking and oil stain on the fabric surface were identified and applied in practice. The entropy threshold segmentation was applied to image processing. Also,The target region was separated from the background area through iterative operation of maximum entropy threshold segmentation. Namely, the fabric defect area was divided from the normal area, and then image was further characterized, the beggar style fabric in real life was also characterized simultaneously. By comparing the two ways above, the pattern of defect fabric used in actual production was determined. This is an effective use of defect detection and its application.
作者
袁小军
陈晓东
邱莉
熊艳平
YUAN Xiaojun;CHEN Xiaodong;QIU Li;XIONG Yanping(College of Textile and Light Industry,Inner Mongolia University of Technology,Huhhot,Inner Mongolia 010080,China)
出处
《毛纺科技》
CAS
北大核心
2019年第3期66-70,共5页
Wool Textile Journal
关键词
疵点检测
最大熵阈值分割
疵点织物应用
疵点特征匹配
fabric defect detection
maximum entropy threshold algorithm
practical application of defect fabric
defect feature matching
