纺织面料经柳疵病等级的计算机自动鉴别研究

Auto Identification Study on Warp Streaky Defect Grades of Textile Fabric

尝试了实现对织物经柳疵点的计算机客观鉴别。以涤纶经二重织物的布面经柳疵点为研究对象,利用计算机对不同严重程度的经柳布样进行视觉化、数据化处理。通过提取图片像素灰度值,并经观察和统计分析,得出图片列间像素的灰度值具有明显的统计分析意义。同时结合经柳的主观评判,建立了一套客观量化经柳疵点的评判方法及指标,提出了一种新的研究思路。

The computer objective identification of warp streaky defect on fabric was tried to realize in the paper. The warp streaky defect of warp-double fabric woven by the polyester filament was taken as sample. Depending on the computer, many specimens with different degree of defect were transferred into sense of vision and data processing. The gray-scale data were abstracted from pictures pixel. Through observing and statistical analysis, the results show that the statistical variance values of pixel plot gray-scale between rows have obviously important meaning. The paper built an objective evaluation method and index of warp streaky defect by combining with subjective judgment. At the same time, a new idea was proposed.