Fraunhofer-Gesellschaft

Publica

Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

SMART - system for segmentation, matching and reconstruction

 
: Hildebrand, A.; Köthe, U.

Clark, B.P.; Douglas, A. ; Society of Photo-Optical Instrumentation Engineers -SPIE-, Bellingham/Wash.:
State-of-the-Art Mapping : 13-15 April 1993, Orlando, Florida
Bellingham/Wash.: SPIE, 1993 (SPIE Proceedings Series 1943)
ISBN: 0-8194-1179-5
pp.66-78
State-of-the-Art Mapping Conference <1993, Orlando/Fla.>
English
Conference Paper
Fraunhofer IGD ()
3D; computer; computing; matching; motion; Rekonstruktion; Segmentierung; shape; vision; visual

Abstract
In many areas of application, such as medicine, robot technology, photogrammetry Ä9Ü etc., the acquisition of an abstract description of three dimensional objects is an important task. A common approach to this problem are the photogrammetric methods. Although the basic algorithms within this field are well known, many questions are still open. Among other problems these methods require an exact determination of the camera positions before the photos can be taken. Additionally the measurement of points in the images and the combination of data from different views often has to be done bz hand. Therefore a lot of skilled work and specialized equipment is necessary during both the acquisition and the evaluation of an image series. Our approach is directed to an integrated system called SMART (Segmentation Matching And ReconsTruction) Ä7Ü Ä8Ü Ä10Ü that is based on general purpose equipment (general purpose workstation, photographic camera). It is designed as a self-calibrating system, i.e . the camera positions, as well as their relative orientations, are derived automatically during the evaluation of the image series. Hence the photos need not to be taken by a specially trained person. The whole procedure within the SMART syste can be reflected in a vision pipeline (see section). After the image acquisition we perform a rough segmentation of the images (1) to find characteristic geometric primitives of the objects and to reject non characteristic ones those occurrence is unavoidable during the acquisition process and (2) to calculate the camera positions approximately. Based on this information we measure the exact positions of the characteristic details and their correspondence. This data allows the usage of an self-calibrating reconstruction algorithm. Now, the obtained partial reconstructions are connected to one complete reconstruction. At the same time the precision of the 3D coordinates is improved by means of a bundle block adjustment Ä25Ü Ä14Ü. Since the obtai n

: http://publica.fraunhofer.de/documents/PX-33980.html