Abstract: To find the type of fabric which is easy to produce defects in the production process, and give a feedback to the production process to improve the quality of the fabric, an algorithm based on local binary pattern and Tamura texture features which combined for fabric defect classification is proposed. The main task of the algorithm is to extract feature vectors fabric, local binary pattern describes the texture feature changing from the local, while Tamura texture features method describes the texture feature changing from the global, a good description of defect can be got by combining the two methods. After extracting feature vector, it uses the conjugate gradient BP algorithm to handle the feature vectors. The convergence of conjugate gradient BP algorithm is better, and it improves the training speed and training accuracy. The experimental results show that the proposed algorithm for defect classification has higher accuracy.

Key words: local binary pattern, fabric, Tamura texture features, conjugate gradient BP algorithm

摘要： 为了找出织物在生产过程中易产生疵点的类型，并反馈到生产工序中以提高织物质量，提出一种基于局部二进制模式与Tamura纹理特征方法相结合的织物疵点分类算法。该算法主要完成的任务是对织物特征向量的提取，局部二进制模式从局部或像素邻域描述纹理的特征，Tamura纹理特征方法从全局描述疵点纹理特征，两者结合能更好地描述疵点纹理特征。完成特征向量提取后，选用共轭梯度BP算法来处理特征向量。共轭梯度BP算法收敛性较好，提高了训练速度和训练精度。实验结果表明，提出的算法对疵点分类具有较高的分类准确率。

关键词: 局部二进制模式, 织物, Tamura纹理特征, 共轭梯度BP算法