Robust and Efficient 3D Recognition by Alignment

Title: en_US Robust and Efficient 3D Recognition by Alignment
Creator: en_US Alter, Tao Daniel
Date: 2004-10-20T19:55:26Z
Date Available: 2004-10-20T19:55:26Z
Date Issued: en_US 1992-09-01
Identifier: en_US AITR-1410
uri: http://hdl.handle.net/1721.1/6799
Abstract: en_US Alignment is a prevalent approach for recognizing 3D objects in 2D images. A major problem with current implementations is how to robustly handle errors that propagate from uncertainties in the locations of image features. This thesis gives a technique for bounding these errors. The technique makes use of a new solution to the problem of recovering 3D pose from three matching point pairs under weak-perspective projection. Furthermore, the error bounds are used to demonstrate that using line segments for features instead of points significantly reduces the false positive rate, to the extent that alignment can remain reliable even in cluttered scenes.
Extent: en_US 113 p.; 903052 bytes; 1830006 bytes
Format: application/octet-stream; application/pdf
Language: en_US
Relation: en_US AITR-1410
Subject: en_US computer vision; en_US object recognition; en_US error models; en_US salignment; en_US weak perspective; en_US pose estimation