Efficient human detection in crowded environment based on motion and appearance information

Chuan Shen Hu, Min Chun Hu, Wen-Huang Cheng, Ja Ling Wu

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Detecting human in crowded environment is profitable but challenging in video surveillance. We propose an efficient human detection method by combining both motion and appearance clues. Moving pixels are first extracted by background subtraction, and then a filtering step is used to narrow the range for human template matching. We utilize integral images to fast generate shape information from edge maps of each frame and define the matching probability to be capable of detecting both full-body and partial-body. Representative human templates are constructed by sparse contours on the basis of the point distribution model (PDM). Moreover, linear regression analysis is also applied to adaptively adjust the template sizes. With the aid of the proposed foreground ratio filtering and the multi-sized template matching techniques, our method not only can efficiently detect human in a crowded environment but also largely enhance the detection accuracy.

Original languageEnglish
Title of host publicationICIMCS 2013 - Proceedings of the 5th International Conference on Internet Multimedia Computing and Service
Pages97-100
Number of pages4
DOIs
StatePublished - 16 Sep 2013
Event5th International Conference on Internet Multimedia Computing and Service, ICIMCS 2013 - Huangshan, China
Duration: 17 Aug 201319 Aug 2013

Publication series

NameACM International Conference Proceeding Series

Conference

Conference5th International Conference on Internet Multimedia Computing and Service, ICIMCS 2013
CountryChina
CityHuangshan
Period17/08/1319/08/13

Keywords

  • human detection
  • sparse human contour
  • template matching

Fingerprint Dive into the research topics of 'Efficient human detection in crowded environment based on motion and appearance information'. Together they form a unique fingerprint.

Cite this