| 融合LBP与GLCM的人群密度分类算法 |
| |
| 引用本文： | 薛翠红,于洋,张朝,杨鹏,李扬.融合LBP与GLCM的人群密度分类算法[J].电视技术,2015,39(24):7-10. |
| |
| 作者姓名： | 薛翠红 于洋 张朝 杨鹏 李扬 |
| |
| 作者单位： | 河北工业大学 控制科学与工程学院,河北工业大学 计算机科学与软件学院,河北工业大学 计算机科学与软件学院,河北工业大学 控制科学与工程学院,中国联合网络通信有限公司天津市分公司 |
| |
| 基金项目： | 天津市科技计划项目；河北省自然科学基金面上项目 |
| |
| 摘 要： | 针对中高密度人群图像检测分类对公共安全的重要性，提出融合局部二值模式LBP与灰度共生矩阵GLCM特征提取的人群密度分类方法。首先用旋转不变的LBP算子进行滤波，得到LBP图像，然后提取滤波后图像的GLCM特征，这样既可以避免LBP算子特征降维带来的损失，又能充分利用LBP和GLCM纹理特征提取的有效性，最后采用有向无环图支持向量机DAGSVM进行密度分类。在Pets2009基准数据库中的实验结果显示该算法具有较高的准确率。
|
| 关 键 词： | 人群密度检测 纹理特征 LBP GLCM DAGSVM |
| 收稿时间： | 2015/5/12 0:00:00 |
| 修稿时间： | 2015/6/23 0:00:00 |
Fusing LBP and GLCM for Crowd Density Classification Algorithm |
| |
| XUE Cui-hong,YU Yang,ZHANG Zhao,YANG Peng and YI Yang.Fusing LBP and GLCM for Crowd Density Classification Algorithm[J].Tv Engineering,2015,39(24):7-10. |
| |
| Authors: | XUE Cui-hong YU Yang ZHANG Zhao YANG Peng and YI Yang |
| |
| Affiliation: | School of Control SScience and Engineering,Hebei University of Technology,School of Computer Science and Engineering,Hebei University of Technology,School of Computer Science and Engineering,Hebei University of Technology,School of Control Science and Engineering,Hebei University of Technology,China United Network Communications Corporation Limited |
| |
| Abstract: | According to the importance of high crowd density image detection and classification for public security, a crowd density classification method is proposed which fuses LBP and GLCM model. Firstly, the rotation-invariant LBP operator is used to filter images to get LBP maps. Then, the GLCM features are extracted in LBP images which can avoid the loss of dimensionality reduction and take advantage of LBP and GLCM in feature extraction. Finally, the Directed Acyclic Graph Support Vector Machines (DAGSVM) is used to make classification. The experimental results in Pets2009 image database show that the proposed algorithm has a good accuracy in crowd density classification. |
| |
| Keywords: | Crowd Density Detection Texture Analysis LBP GLCM DAGSVM |
| 本文献已被 万方数据 等数据库收录！ |
| 点击此处可从《电视技术》浏览原始摘要信息 |
|
点击此处可从《电视技术》下载全文 |
|
