经编布匹瑕疵点检测方法

Method for detecting defects in warp knitted fabric

与梭织纬编布匹不同,经编布匹由于编织工艺的原因造成布匹网纹稀疏、粗大,导致现有的视觉疵点检测方法只能针对少数种类的经编布匹上的疵点有效果,普适性不佳.针对经编布匹的特点,提出了一种新的视觉检测方法.该方法基于最大熵阈值分割和连通域的思想,有效解决了经编布匹疵点检测中的误检问题.通过对上百张样本照片、数十种疵点种类进行了试验,结果表明该方法具有较好的准确性和普适性,可以完成对数十种经编布匹疵点的检测和定位.

Different from woven weft-knitted fabric,warp-knitted fabric had sparse and bulky network lines due to weaving technology,which resulted in the existing visual detection methods for defects,which could only be used for a few kinds of warp-knitted fabric.According to the characteristics of warp-knitted fabric,a new visual detection method was proposed in this paper.Based on the idea of maximum entropy threshold segmentation and connected domain,this method effectively solved the problem of false detection in warp-knitted fabric defect detection.Through the test of hundreds of sample photos and dozens of defect types,the results showed that this method had good accuracy and universality,and could complete the detection and localization of dozens of warp-knitted fabric defects.