基于MBLBPV算法的布匹瑕疵检测方法

针对布匹瑕疵检测，在传统局部二值模式（Local Binary Pattern，LBP）与局部二值模式方差（LBP Variance，LBPV）的基础上，提出一种基于多尺度分块局部二值模式方差（Multi Scale Block Local Binary Patterns Variance,MBLBPV）的检测算法。首先，采用适当尺度大小的子区域灰度均值代替单像素灰度值，提取LBP特征，以降低噪声影响；然后，融合图像区域对比度信息，并将其作为编码值的权重，提取图像MBLBPV特征，并基于该特征实现瑕疵的检测。实验结果表明，相对于传统方法，MBLBPV抗噪力强、检测正确率更高。

Fabric Defect Detection Based on Multi-Scale Block Local Binary Patterns Variance

Based on the truditional local binary pattern (LBP) and the LBP variance (LBPV),a novel algorithm called multi-scale local binary patterns variance (MBLBPV) is presented for fabric defect detection.Firstly,the average gray level of a subarea of the appropriate scale is used to replace the gray level of a single pixel and the LBP feature is extracted to reduce the noise influence.Then,the image region contrast information is fused and used as the weight of the encoded value to extract the MBLBPV features,and the fabric defects are detected based on the extracted features.Experimental results show that MBLBPV has higher detection accuracy and is more robust to image noise than the traditional approaches.