Pedestrian Counting in Video Sequences based on Optical Flow Clustering

International Journal of Image Processing

Sizuka Fujisawa¹  Hirotaka Nakano¹  Go Hasegawa¹  Yoshiaki Taniguchi¹ 

[1] $$

关键词: Pedestrian Counting;    Video Processing;    Optical Flow;    Clustering.;   

DOI  :  

来源: Computer Science Journals

【 摘 要 】

The demand for automatic counting of pedestrians at event sites, buildings, or streets has beenincreased. Existing systems for counting pedestrians in video sequences have a problem thatcounting accuracy degrades when many pedestrians coexist and occlusion occurs frequently. Inthis paper, we introduce a method of clustering optical flows extracted from pedestrians in videoframes to improve the counting accuracy. The proposed method counts the number ofpedestrians by using pre-learned statistics, based on the strong correlation between the numberof optical flow clusters and the actual number of pedestrians. We evaluate the accuracy of theproposed method using several video sequences, focusing in particular on the effect ofparameters for optical flow clustering. We find that the proposed method improves the countingaccuracy by up to 25% as compared with a non-clustering method. We also report that using aclustering threshold of angles less than 1 degree is effective for enhancing counting accuracy.Furthermore, we compare the performance of two algorithms that use feature points and latticepoints when optical flows are detected. We confirm that the counting accuracy using featurepoints is higher than that using lattice points especially when the number of occluded pedestriansincreases.

【 授权许可】

Unknown