图像处理在机织物组织结构识别中的应用

Application of Image Processing in Identification of Woven Fabric Structure

组织结构是机织物的重要规格参数,目前纺织企业主要靠人工分析来完成这一识别,人工识别不但效率低、精度差,还易受主观因素影响。图像处理具有快速、准确、自动化程度高等优点,用来识别织物组织结构可以大大提高检测效率和客观准确度。文章详细阐述了国内外在应用图像处理技术进行机织物组织结构自动识别的研究现状,并将这些识别方法概括为两类:基于时域的方法和基于频域的方法;并总结了研究存在不足的几个方面;最后指出纺织领域的许多人工检测项目,均可用图像处理来代替,图像处理技术在纺织品检测领域具有广阔的应用前景。

Structure is a very important specification parameter for woven fabric. Nowadays, textile enterprises mainly depend on manual analysis to identify woven fabric structure. Manual identification suffers low efficient and poor accuracy, and is easily affected by subjective factors. The image processing technology has such advantages as high speed, accuracy and high degree of automation, so it can be used to identify woven fabric structure, which can increase greatly the efficiency, objectivity and accuracy. This paper elaborates domestic and overseas research status on application of image process technology in automatic identification of woven fabric structure. The identification methods are classified into two groups： methods based on time domain and those based on frequency domain. Then the shortcomings in the researches are summarized. In the end, this paper points out that many manual detection items in the textile field can be replaced by image processing and that the image processing technology owns wide application prospect in textile detection field.