Presentation 2010-03-15
Target identification for multi images using a search tree
Takumi HONMA, Nobuyuki TAKAHASHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) We develop a target identification that is robust in a variety of changed situation and noises uses search tree and show availability of search tree in target identification by simulation with prior information. Target chase method that used multi cameras has been researched in the past research. But, accuracy decreases by the situation of each image. For example, the character of the target, change in background, and mixing of noises. And so we develop a new target identification method that applies Monte Carlo tree search. Monte Carlo tree search is used in search problem in game theory. The tree search take out various features from images. Therefore, this method can detect and track a targets including obstacles that single-angle can't track, specific of a target detected from one angle that cannot be discriminated and information extraction from image databases. Also we used statistical characteristics for not perfect matching images. Therefore, not perfect matching images that include occlusion and noise can be thought the same target. This search method can do high adaptability target identification that uses the best feature.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Target identicication / Multi-images / Search tree / Statistical characteristics
Paper # IE2009-182
Date of Issue

Conference Information
Committee IE
Conference Date 2010/3/8(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 Image Engineering (IE)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Target identification for multi images using a search tree
Sub Title (in English)
Keyword(1) Target identicication
Keyword(2) Multi-images
Keyword(3) Search tree
Keyword(4) Statistical characteristics
1st Author's Name Takumi HONMA
1st Author's Affiliation Graduate school, FUTURE UNIVERSITY HAKODATE()
2nd Author's Name Nobuyuki TAKAHASHI
2nd Author's Affiliation Graduate school, FUTURE UNIVERSITY HAKODATE
Date 2010-03-15
Paper # IE2009-182
Volume (vol) vol.109
Number (no) 469
Page pp.pp.-
#Pages 6
Date of Issue