基于MB-LBP和HOG的掌纹识别

发布时间：2018-06-13 14:28

本文选题：掌纹识别 + 多块二值模式　；参考：《计算机应用研究》2017年03期

【摘要】：掌纹识别是模式识别及智能信息处理领域的研究热点。针对传统掌纹识别方法易受噪声影响,且旋转鲁棒性差的问题,提出基于分区的多块局部二值模式(MB-LBP)和梯度方向直方图(HOG)的掌纹识别方法。该算法利用分区MB-LBP和HOG算法分别提取掌纹的纹理、边缘特征,将提取的两种特征通过串联的方式进行融合,最后计算测试图像与训练图像的绝对值距离并利用最近邻分类器得出分类结果。实验对比传统算法(PCA、LBP和HOG),得到较高的掌纹识别率。因此,将基于分区多块局部二值模式与梯度方向直方图特征进行融合,可提取较完整的掌纹有效信息,对于光照不均匀和有位置变化的掌纹具有一定的健壮性,具有良好的掌纹识别性能。
[Abstract]:Palmprint recognition is a hot topic in the field of pattern recognition and intelligent information processing. Aiming at the problem that traditional palmprint recognition methods are susceptible to noise and poor rotation robustness, a method of palmprint recognition based on multi-block local binary mode MB-LBPand gradient direction histogram (hog) is proposed. The algorithm uses partitioned MB-LBP and Hog algorithms to extract palmprint texture and edge features, and combines the two features in series. Finally, the absolute distance between the test image and the training image is calculated and the classification results are obtained by using the nearest neighbor classifier. The experimental results show that the recognition rate of palmprint is higher than that of PCA-LBP and hog. Therefore, based on multi-block local binary pattern and gradient direction histogram feature fusion, more complete palmprint effective information can be extracted, which is robust to the palm print with uneven illumination and position change. Good palmprint recognition performance.
【作者单位】：内蒙古农业大学计算机与信息工程学院;
【基金】：国家自然科学基金资助项目(61562067) 中国博士后科学基金资助项目(20100480370,201104179) 内蒙古自治区高等学校科学研究项目(NJZY068) 内蒙古农业大学基金项目(JCYJ201201,NDPYTD 210-9)
【分类号】：TP391.41

【相似文献】