Visible to the public A fast human detection algorithm of video surveillance in emergencies

TitleA fast human detection algorithm of video surveillance in emergencies
Publication TypeConference Paper
Year of Publication2014
AuthorsMa Juan, Hu Rongchun, Li Jian
Conference NameControl and Decision Conference (2014 CCDC), The 26th Chinese
Date PublishedMay
Keywordsbackground subtraction, Conferences, detection algorithms, dilation, edge detection, edge tracking of head, Educational institutions, Electronic mail, emergencies, Estimation, fame subtraction, frame subtraction, Gaussian filter, Gaussian processes, human detection algorithm, IEEE Computer Society, Image edge detection, moving target, object detection, single Guassian model, target mask, video surveillance
Abstract

This paper propose a fast human detection algorithm of video surveillance in emergencies. Firstly through the background subtraction based on the single Guassian model and frame subtraction, we get the target mask which is optimized by Gaussian filter and dilation. Then the interest points of head is obtained from figures with target mask and edge detection. Finally according to detecting these pionts we can track the head and count the number of people with the frequence of moving target at the same place. Simulation results show that the algorithm can detect the moving object quickly and accurately.

URLhttps://ieeexplore.ieee.org/document/6852404/
DOI10.1109/CCDC.2014.6852404
Citation Key6852404