An Improved Moving Multi-Human Target Detection Algorithm

Liang Feng-Mei, Tong Lin-Lin

Abstract


In the detection of moving multi-human targets, the major problems existing lie in the detection speed and precision. Fortunately, the HOG feature presents a very considerable effect on the detection accuracy. However, the problem of low detecting speed caused by its large amount of calculation prevents the HOG feature from being well applied in scenes where the real-time requirements are needed. Given this problem, this paper presents a method which combines the Gaussian mixture background model and HOG feature. This method solved firstly by the Gaussian mixture background model to detect the moving foreground in the video. And then use HOG+SVM to handle the moving foreground that has been detected. As a result, the amount of computation is reduced considerably and the real-time performance of the HOG algorithm is improved greatly. Verified by the experiment, the detection accuracy of this algorithm can reach 94%.

 

DOI: http://dx.doi.org/10.11591/telkomnika.v11i12.3747


Keywords


Multi-human targets; The Gaussian mixture background model; HOG feature

Full Text:

PDF

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Indonesian Journal of Electrical Engineering and Computer Science (IJEECS)
p-ISSN: 2502-4752, e-ISSN: 2502-4760
This journal is published by the Institute of Advanced Engineering and Science (IAES) in collaboration with Intelektual Pustaka Media Utama (IPMU).

shopify stats IJEECS visitor statistics