Tools

LEADER 03472nam a2200697Ia 4500

001 ocn227911144;

003 OCoLC;

005 20141212054704.0;

006 m d f ;

007 cr mn|||||||||;

008 080513s2006 mdua ob f000 0 eng d;

029 0 a| DTICEb| ADA459273;

035 a| (Sirsi) o227911144;

035 a| (OCoLC)227911144;

040 a| DTICEc| DTICEd| GPOd| UtOrBLW;

043 a| n-us---;

049 a| ERE#;

074 a| 0324-A-01 (online);

086 0 a| D 101.133:4006;

100 1 a| David, Philip.=| ^A25388;

245 10 a| Shaped-based recognition of 3D objects from 2D projections /c| by Philip David and Daniel DeMenthon.;

246 3 a| Shaped based recognition of 3D objects from 2D projections;

260 a| Adelphi, MD :b| Army Research Laboratory,c| [2006];

300 a| 1 online resource (vi, 32 pages) :b| illustrations.;

336 a| text2| rdacontent;

337 a| computer2| rdamedia;

338 a| online resource2| rdacarrier;

490 1 a| ARL-TR ;v| 4006;

500 a| Title from PDF title screen (ARL, viewed Nov. 22, 2010).;

500 a| "December 2006.";

500 a| Prepared in collaboration with University of Maryland, College Park, MD.;

504 a| Includes bibliographical references.;

506 a| APPROVED FOR PUBLIC RELEASE.;

513 a| Summary.b| Jan. to July 2006.;

520 a| We present an object recognition algorithm that uses model and image line features to locate complex objects in high clutter environments. Corresponding line features are determined by a three-stage process. The first stage generates a large number of approximate pose hypotheses from correspondence of one or two lines in the model and image. Next, pose hypotheses from the previous stage are quickly evaluated and ranked by a comparison of local image neighborhoods to the corresponding local model neighborhoods. Fast nearest neighbor and range search algorithms are used to implement a distance measure that is unaffected by clutter and partial occlusion. The ranking of pose hypotheses is invariant to changes in image scale, orientation, and partially invariant to affine distortion. Finally, a robust pose estimation algorithm is applied for refinement and verification, starting from the few best approximate poses produced by the previous stages. Experiments on real images demonstrate robost recognition of partially occluded objects in very high clutter environments.;

650 0 a| Cybernetics.=| ^A8612;

650 0 a| Algorithms.=| ^A17554;

650 17 a| Cybernetics.2| scgdst;

650 17 a| Algorithms.2| dtict;

650 17 a| Clutter.2| dtict;

650 17 a| Recognition.2| dtict;

650 27 a| Image registration.2| dtict;

650 27 a| Lines(geometry)2| dtict;

653 1 a| OBJECT RECOGNITION;

653 2 a| SOFT POSIT;

653 2 a| SHAPE FEATURES;

653 2 a| POSE HYPOTHESES;

653 2 a| LINE NEIGHBORHOODS;

653 2 a| LINE DETECTION;

653 2 a| AFFINE TRANSFORMATION;

653 2 a| FEATURE CORRESPONDENCE ALGORITHMS;

700 1 a| DeMenthon, Daniel.=| ^A1028399;

710 2 a| U.S. Army Research Laboratory.=| ^A367731;

830 0 a| ARL-TR (Aberdeen Proving Ground, Md.) ;v| 4006.=| ^A566074;

856 40 u| https://purl.fdlp.gov/GPO/LPS125451;

949 a| Click on web addressw| ASISh| JOYNER94;

994 a| C0b| ERE;

998 a| 2390193;

998 a| u2390193;

596 a| 1;

999 a| CLICK ON WEB ADDRESSw| ASISc| 1i| 2390193-1001l| JNETm| JOYNERr| Ys| Yt| JNE1DOu| 1/11/2011x| EDOCUMENTz| JERESOURCE;

Shaped-based recognition of 3D objects from 2D projections / by Philip David and Daniel DeMenthon.

Author/creator	David, Philip
Other author	DeMenthon, Daniel.
Other author	U.S. Army Research Laboratory.
Format	Electronic
Publication Info	Adelphi, MD : Army Research Laboratory, [2006]
Description	1 online resource (vi, 32 pages) : illustrations.
Supplemental Content	https://purl.fdlp.gov/GPO/LPS125451
Subjects	Cybernetics. Algorithms.

Variant title	Shaped based recognition of 3D objects from 2D projections
Series	ARL-TR ; 4006 ARL-TR (Aberdeen Proving Ground, Md.) ; 4006. ^A566074
Abstract	We present an object recognition algorithm that uses model and image line features to locate complex objects in high clutter environments. Corresponding line features are determined by a three-stage process. The first stage generates a large number of approximate pose hypotheses from correspondence of one or two lines in the model and image. Next, pose hypotheses from the previous stage are quickly evaluated and ranked by a comparison of local image neighborhoods to the corresponding local model neighborhoods. Fast nearest neighbor and range search algorithms are used to implement a distance measure that is unaffected by clutter and partial occlusion. The ranking of pose hypotheses is invariant to changes in image scale, orientation, and partially invariant to affine distortion. Finally, a robust pose estimation algorithm is applied for refinement and verification, starting from the few best approximate poses produced by the previous stages. Experiments on real images demonstrate robost recognition of partially occluded objects in very high clutter environments.
General note	Title from PDF title screen (ARL, viewed Nov. 22, 2010).
General note	"December 2006."
General note	Prepared in collaboration with University of Maryland, College Park, MD.
Bibliography note	Includes bibliographical references.
Access restriction	APPROVED FOR PUBLIC RELEASE.
Report note	Summary. Jan. to July 2006.
GPO item number	0324-A-01 (online)
Govt. docs number	D 101.133:4006

Availability

Library	Location	Call Number	Status	Item Actions
	Electronic Resources	Access Content Online	✔ Available

Shaped-based recognition of 3D objects from 2D projections / by Philip David and Daniel DeMenthon.

Click here for more information about this title

Availability