南開科技大學圖書館 |

Language: 繁體中文

說明(常見問題)

南開科技大學

圖書館首頁

編目中圖書申請

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

Multi-resolution adaptive object rec...

Levner, Ilya.

Multi-resolution adaptive object recognition system: A step towards autonomous vision systems.

紀錄類型:	書目-電子資源 : 單行本
正題名/作者:	Multi-resolution adaptive object recognition system: A step towards autonomous vision systems./
作者:	Levner, Ilya.
面頁冊數:	102 p.
附註:	Source: Masters Abstracts International, Volume: 42-05, page: 1758.
Contained By:	Masters Abstracts International42-05.
標題:	Computer Science. -
電子資源:	Download fulltext (下載全文)
ISBN:	0612878074

Multi-resolution adaptive object recognition system: A step towards autonomous vision systems.
Levner, Ilya.

Multi-resolution adaptive object recognition system: A step towards autonomous vision systems. - 102 p.

Source: Masters Abstracts International, Volume: 42-05, page: 1758.

Thesis (M.Sc.)--University of Alberta (Canada), 2003.

Modern attempts at robust object recognition attempted to learn adaptive strategies. One such system was ADORE (Adaptive Object Recognition). The system successfully learned dynamic object recognition strategies for finding buildings in aerial images. By modelling object recognition as a Markov Decision Process, where the intermediate representations are continuous state spaces, and vision procedures are actions, ADORE was able to select the next action (i.e., vision routine) at each step so as to maximize the quality of the final image interpretation. In contrast to the majority of vision systems, ADORE did not employ a static sequence of vision routines but adaptively selected the most appropriate sequence based on image content.

ISBN: 0612878074Subjects--Topical Terms:

1000005419
Computer Science.

Multi-resolution adaptive object recognition system: A step towards autonomous vision systems.
LDR:02133nmm 2200289 4500 001 1000004347
005 20051102155248.5
008 141201s2003 eng d
020 $a 0612878074
035 $a (UnM)AAIMQ87807
035 $a AAIMQ87807
040 $a UnM $c UnM{me_controlnum}
100 1 $a Levner, Ilya. $3 1000005710
245 1 0 $a Multi-resolution adaptive object recognition system: A step towards autonomous vision systems.
300 $a 102 p.
500 $a Source: Masters Abstracts International, Volume: 42-05, page: 1758.
500 $a Adviser: Vadim Bulitko.
502 $a Thesis (M.Sc.)--University of Alberta (Canada), 2003.
520 $a Modern attempts at robust object recognition attempted to learn adaptive strategies. One such system was ADORE (Adaptive Object Recognition). The system successfully learned dynamic object recognition strategies for finding buildings in aerial images. By modelling object recognition as a Markov Decision Process, where the intermediate representations are continuous state spaces, and vision procedures are actions, ADORE was able to select the next action (i.e., vision routine) at each step so as to maximize the quality of the final image interpretation. In contrast to the majority of vision systems, ADORE did not employ a static sequence of vision routines but adaptively selected the most appropriate sequence based on image content.
520 $a As a pioneering concept, ADORE demonstrated the applicability and effectiveness of AI and machine learning techniques at automatically building vision systems. This thesis continues the research into automatic creation of object recognition systems, by investigating methods and techniques that minimize the need for human intervention while maximizing the performance and portability of a computer vision system. (Abstract shortened by UMI.)
590 $a School code: 0351.
650 4 $a Computer Science. $3 1000005419
650 4 $a Artificial Intelligence. $3 165300
690 $a 0984
690 $a 0800
710 2 0 $a University of Alberta (Canada). $3 1000005530
773 0 $t Masters Abstracts International $g 42-05.
790 1 0 $a Bulitko, Vadim, $e advisor
790 $a 0351
791 $a M.Sc.
792 $a 2003
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=MQ87807 $z Download fulltext (下載全文)