東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Efficient selection of disambiguatin...

Schaeffer, Monika.

Linked to FindBook

Google Book

Amazon

博客來

Efficient selection of disambiguating actions for stereo vision.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Efficient selection of disambiguating actions for stereo vision./
Author:	Schaeffer, Monika.
Description:	57 p.
Notes:	Source: Masters Abstracts International, Volume: 49-01, page: 0481.
Contained By:	Masters Abstracts International49-01.
Subject:	Engineering, Robotics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1479851
ISBN:	9781124159263

Efficient selection of disambiguating actions for stereo vision.
Schaeffer, Monika.

Efficient selection of disambiguating actions for stereo vision. - 57 p.

Source: Masters Abstracts International, Volume: 49-01, page: 0481.

Thesis (M.S.)--Duke University, 2010.

In many domains that involve the use of sensors, such as robotics or sensor networks, there are opportunities to use some form of active sensing to disambiguate data from noisy or unreliable sensors. These disambiguating actions typically take time and expend energy. One way to choose the next disambiguating action is to select the action with the greatest expected entropy reduction, or information gain. In this work, we consider active sensing in aid of stereo vision for robotics. Stereo vision is a powerful sensing technique for mobile robots, but it can fail in scenes that lack strong texture. In such cases, a structured light source, such as vertical laser line, can be used for disambiguation. By treating the stereo matching problem as a specially structured HMM-like graphical model, we demonstrate that for a scan line with n columns and maximum stereo disparity d, the entropy minimizing aim point for the laser can be selected in O( nd) time - cost no greater than the stereo algorithm itself. A typical HMM formulation would suggest at least O(nd 2) time for the entropy calculation alone. We demonstrate the effectiveness of this approach on some artificial data sets. We also present results on a prototype hybrid stereo/laser device of our own construction. We demonstrate that this device can produce good disparity maps of realistic scenes of the type that a mobile robot might encounter, and that our algorithm is effective on the data generated by our prototype device.

ISBN: 9781124159263Subjects--Topical Terms:

1018454
Engineering, Robotics.

Efficient selection of disambiguating actions for stereo vision.
LDR:02440nam 2200313 4500 001 1399343
005 20110928091744.5
008 130515s2010 ||||||||||||||||| ||eng d
020 $a 9781124159263
035 $a (UMI)AAI1479851
035 $a AAI1479851
040 $a UMI $c UMI
100 1 $a Schaeffer, Monika. $3 1678298
245 1 0 $a Efficient selection of disambiguating actions for stereo vision.
300 $a 57 p.
500 $a Source: Masters Abstracts International, Volume: 49-01, page: 0481.
500 $a Adviser: Ron Parr.
502 $a Thesis (M.S.)--Duke University, 2010.
520 $a In many domains that involve the use of sensors, such as robotics or sensor networks, there are opportunities to use some form of active sensing to disambiguate data from noisy or unreliable sensors. These disambiguating actions typically take time and expend energy. One way to choose the next disambiguating action is to select the action with the greatest expected entropy reduction, or information gain. In this work, we consider active sensing in aid of stereo vision for robotics. Stereo vision is a powerful sensing technique for mobile robots, but it can fail in scenes that lack strong texture. In such cases, a structured light source, such as vertical laser line, can be used for disambiguation. By treating the stereo matching problem as a specially structured HMM-like graphical model, we demonstrate that for a scan line with n columns and maximum stereo disparity d, the entropy minimizing aim point for the laser can be selected in O( nd) time - cost no greater than the stereo algorithm itself. A typical HMM formulation would suggest at least O(nd 2) time for the entropy calculation alone. We demonstrate the effectiveness of this approach on some artificial data sets. We also present results on a prototype hybrid stereo/laser device of our own construction. We demonstrate that this device can produce good disparity maps of realistic scenes of the type that a mobile robot might encounter, and that our algorithm is effective on the data generated by our prototype device.
590 $a School code: 0066.
650 4 $a Engineering, Robotics. $3 1018454
650 4 $a Artificial Intelligence. $3 769149
650 4 $a Computer Science. $3 626642
690 $a 0771
690 $a 0800
690 $a 0984
710 2 $a Duke University. $b Computer Science. $3 1019283
773 0 $t Masters Abstracts International $g 49-01.
790 1 0 $a Parr, Ron, $e advisor
790 1 0 $a Tomasi, Carlo $e committee member
790 1 0 $a Welch, Greg $e committee member
790 $a 0066
791 $a M.S.
792 $a 2010
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=1479851