计算机工程与应用 ›› 2008, Vol. 44 ›› Issue (28): 169-171.DOI: 10.3778/j.issn.1002-8331.2008.28.056

• 图形、图像、模式识别 • 上一篇    下一篇

基于灰度共生矩阵的密集人群人数估计

郭 森1,严和平2,柳 伟1   

  1. 1.深圳信息职业技术学院,广东 深圳 518029
    2.中国人民解放军75771部队,广州 510540
  • 收稿日期:2008-05-30 修回日期:2008-07-25 出版日期:2008-10-01 发布日期:2008-10-01
  • 通讯作者: 郭 森

Crowd density estimation based on Grey Level Dependence Matrix

GUO Sen1,YAN He-ping2,LIU Wei1   

  1. 1.Shenzhen Institute of Information Technology,Shenzhen,Guangdong 518029,China
    2.No 75771 Unit of PLA,Guangzhou 510540,China
  • Received:2008-05-30 Revised:2008-07-25 Online:2008-10-01 Published:2008-10-01
  • Contact: GUO Sen

摘要: 从统计学的角度研究了开阔场所中密集人群估计的方法。首先基于场景图像的灰度共生矩阵提取16个特征,构建训练样本的特征矩阵。然后应用主成份分析方法找出其中最为有效的几项特征指标,利用回归分析的方法找出特征指标与人群数量之间的相关关系,建立统计模型,进而运用该模型估计场景图像中的人群数量。通过实验证明了该方法的有效性。

关键词: 人数估计, 密集人群, 灰度共生矩阵

Abstract: This paper discusses people counting of crowd open scene according to statistics.Firstly,16 features of texture of scene image is obtained based on grey level dependence matrix which is called feature vector,and feature matrix is constructed based on those vectors.Then the most significant features are taken based on principal component analysis and the statistical model of number of people which are included in image according to feature vector is built.At last,people are counted based on those statistical models.According to the experiments,the results show that this method is efficient and feasible.

Key words: people counting, crowd scene, grey level dependence matrix