织物疵点自动识别技术在毛纺中的应用

为了实现自动验布，摆脱人工验布的种种缺点，介绍了采用神经网络和图像处理技术识别织物疵点的准确性和可行性，把疵点主要分为破洞、断经、断纬、油渍、筘路等5类，使用灰度统计量法对疵点图像进行处理，然后应用BP网络对其进行分类识别。实验证明该方法的准确性高、速度快，符合自动验布系统的要求。同时文章还讨论了该技术在毛纺企业的应用前景。

Application of the technology to indentify fabric defect in wool textile

To get the automation cloth inspection,and to cast off the disadvantages of manual cloth inspection,Introduces using neural network and image handle to identify fabric defect,and dividing the defect into five types:holes,warp-lacking,weft-lacking,oil stains,and buckles.Uses the dot deepness measure to handle image,and uses BP net to identify the data.The test proves that this method is very nicety and very speediness,it fits for the automation cloth inspecting system.And the problems of the practicality of the application are commented on.