DocumentCode :
721088
Title :
Extracting Recurrent Motion Flows from Crowded Scene Videos: A Coherent Motion-Based Approach
Author :
Yang Mi ; Lihang Liu ; Weiyao Lin ; Weiyue Wang
Author_Institution :
Dept. of Electron. Eng., Shanghai Jiao Tong Univ., Shanghai, China
fYear :
2015
fDate :
20-22 April 2015
Firstpage :
371
Lastpage :
376
Abstract :
In this paper, we propose a new approach which utilizes coherent motion regions to extract and visualize recurrent motion flows in crowded scene surveillance videos. The proposed approach first extract coherent motion regions from a crowded scene video. Then a frame-level clustering process is proposed to cluster frames into different recurrent-motion-pattern (RMP) groups according to the coherent-region similarity between frames. By merging similar coherent regions from the same RMP group, we can achieve motion flow regions representing the major motion flows in each recurrent motion pattern. Finally, a flow curve extraction process is also proposed which extracts flow curves from motion flow regions to provide a proper visualization of the recurrent motion patterns. Experimental results demonstrate that our approach can precisely achieve recurrent motion flows for various crowded scene videos.
Keywords :
feature extraction; image matching; image motion analysis; pattern clustering; video signal processing; video surveillance; RMP group; coherent motion region extraction; coherent-region similarity; crowded scene surveillance videos; flow curve extraction process; frame-level clustering process; recurrent motion flow extraction; recurrent motion pattern visualization; recurrent-motion-pattern groups; Feature extraction; High definition video; Merging; Motion segmentation; Semantics; Trajectory; Videos; Coherent motion; Crowded scene; Traffic flow extraction;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Multimedia Big Data (BigMM), 2015 IEEE International Conference on
Conference_Location :
Beijing
Print_ISBN :
978-1-4799-8687-3
Type :
conf
DOI :
10.1109/BigMM.2015.20
Filename :
7153917
Link To Document :
بازگشت