Elastic matching of very large digital images on high performance clusters

J. Modersitzki, W. Obelöer, O. Schmitt, G. Lustig

Abstract

The aim of the human neuroscanning project is to build an atlas of the human brain, based on a variety of image modalities in particular histological sections of a prepared brain. Reconstructing essential information out of deformed images is a key problem. We describe a method to correct elastic deformations. Since the method is computational expensive a parallel implementation is presented. The measurements and results shown are performed on a cluster of 48 Pentium II PCs connected via Myrinet.
Original languageEnglish
Title of host publicationHigh-Performance Computing and Networking
EditorsPeter Sloot, Marian Bubak, Alfons Hoekstra, Bob Hertzberger
Number of pages9
Volume1593
Place of PublicationBerlin, Heidelberg
PublisherSpringer Berlin Heidelberg
Publication date01.04.1999
Pages141-149
ISBN (Print)978-3-540-65821-4
ISBN (Electronic)978-3-540-48933-7
DOIs
Publication statusPublished - 01.04.1999
Event7th International Conference on High-Performance Computing and Networking - Amsterdam, Netherlands
Duration: 12.04.199914.04.1999
Conference number: 150469

Fingerprint

Dive into the research topics of 'Elastic matching of very large digital images on high performance clusters'. Together they form a unique fingerprint.

Cite this