¹Institute of Software and Intelligent Technology, Hangzhou Dianzi University, Hangzhou, 310018, China
²Zhejiang Provincial Engineering Center on Media Data Cloud Processing and Analysis, Hangzhou Dianzi University, Hangzhou, 310018, China
³College of Computer Science and Technology, Zhejiang University, Hangzhou, 310027, China

---

Abstract—Crowd density estimation is very important in intelligent crowd monitoring. In this paper, a new approach of crowed density estimation is proposed. This method combines the advantage of pixel statistical feature and texture analysis, and reduces the impact of perspective distortion by dividing the region of interest. Moreover, we estimate the crowd size (the range of numbers) for high density and extremely high density crowd. The experiment results show that the proposed method has a relatively high accuracy in the whole density range.

Index Terms—Region of interest, Crowd density estimation, Crowd size estimation, Pixel statistical, GLCM.

[PDF]

Cite: Ming Jiang, Jingcheng Huang, Xingqi Wang, Jingfan Tang, Chunming Wu, "An Approach for Crowd Density and Crowd Size Estimation," Journal of Software vol. 9, no. 3, pp. 757-762, 2014.