Presentation 1996/11/22
Detection of moving objects from MPEG video stream
Akio Yoneyama, Yasuyuki Nakajima, Kiyono Ujihara,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Independent processing is usually performed in moving object detection for remote surveillance and video analysis, and in video compression for storage and transmission. Therefore, when video compression is required for storage after detection of moving object, it is necessary to use computationally very heavy motion estimation in each processing, which results in very expensive and inefficient system. In this paper, we propose efficient moving object detection from MPEG coded video data using motion vector information and DCT coefficients. Since this method directly uses coding information, pixel domain processing is not required and therefore very fast and efficient moving object detection is realized.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) MPEG / Motion Vector / DCT / Moving object detection / Remote surveillance / Video Analysis
Paper # PRMU96-101
Date of Issue

Conference Information
Committee PRMU
Conference Date 1996/11/22(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Detection of moving objects from MPEG video stream
Sub Title (in English)
Keyword(1) MPEG
Keyword(2) Motion Vector
Keyword(3) DCT
Keyword(4) Moving object detection
Keyword(5) Remote surveillance
Keyword(6) Video Analysis
1st Author's Name Akio Yoneyama
1st Author's Affiliation KDD Co. Ltd., R&D Laboratories()
2nd Author's Name Yasuyuki Nakajima
2nd Author's Affiliation KDD Co. Ltd., R&D Laboratories
3rd Author's Name Kiyono Ujihara
3rd Author's Affiliation KDD Co. Ltd., R&D Laboratories
Date 1996/11/22
Paper # PRMU96-101
Volume (vol) vol.96
Number (no) 385
Page pp.pp.-
#Pages 8
Date of Issue