基于SMO算法的织物组织结构识别

摘要

提出了一种用机器识别布料结构的方法.该方法采用图像去噪、增强及二值化技术对织物组织图进行预处理,采用经纬像素差值法提取出织物组织结构的特征向量,用序列最小化(SMO)算法进行识别分类,重构出清晰的、便于生产加工的织物组织结构图.实验结果表明,通过该方法对织物组织结构的识别具有较高的准确率.%There is a new method for recognition of fabric organization. Firstly, the image technology is used to preprocess the fabric organization image, and a special way adopted to pick-up the support vector from fabric organization structure. Then, the SMO arithmetic is used to recognize and classify. Lastly, the fabric organization image which facilitates the production and processing is reconstructed. Test results show that this method has a good accuracy on the recognition of fabric organization structure.