Real-Time Extraction of Unknown Region from Dynamic Background by Robust Projection onto Eigen Space

Satoshi KAWABATA  Sinsaku HIURA  Kosuke SATO  

Publication
D - Abstracts of IEICE TRANSACTIONS on Information and Systems (Japanese Edition)   Vol.J89-D   No.4   pp.826-835
Publication Date: 2006/04/01
Online ISSN: 1881-0225
DOI: 
Print ISSN: 1880-4535
Type of Manuscript: PAPER
Category: 
Keyword: 
background subtraction,  dynamic background,  eigen space,  robust projection,  

Full Text(in Japanese): PDF(831.9KB)
>>Buy this Article


Summary: 
In computer vision, the background subtraction method is widely used to extract a changing region in a scene. The method determines the region by comparing an image with the background image obtained in advance. To expand the method for dynamic scene, abundant improvements based on updating a background have been proposed. However, any object satisfies the update conditions may be replaced with original values, and the methods have less warrant for the properness of the background image. In this paper, we propose a real-time extraction method of an unknown object region based on the robust projection onto the eigen space which represents dynamically changing background images.