Parallel alignment of a large number of range images

Takeshi Oishi, Atsushi Nakazawa, Ryo Kurazume, Katsushi Ikeuchi, Ryusuke Sagawa

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

This chapter describes a method for parallel alignment of multiple range images. There are problems of computational time and memory space in aligning a large number of range images simultaneously. We developed a parallel method to address the problems. Searching for corresponding points between two range images is time-consuming and requires considerable memory space when performed independently. However, this process can be preformed in parallel, with each corresponding pair of range images assigned to a node. Because the computation time is approximately proportional to the number of vertices, by assigning the pairs so that the number of vertices computed is equal on each node, the load on each node is effectively distributed. In order to reduce the amount of memory required on each node, a hypergraph that represents the correspondences of range images is created, and heuristic graph partitioning algorithms are applied to determine the optimal assignment of the pairs. Moreover, by rejecting redundant dependencies, it becomes possible to accelerate computation time and reduce the amount of memory required on each node. The method was tested on a 16-processor PC cluster, where it demonstrated high extendibility and improved performance.

Original languageEnglish
Title of host publicationDigitally Archiving Cultural Objects
PublisherSpringer US
Pages109-126
Number of pages18
ISBN (Print)9780387758060
DOIs
Publication statusPublished - Dec 1 2008
Externally publishedYes

Fingerprint

Data storage equipment

All Science Journal Classification (ASJC) codes

  • Computer Science(all)

Cite this

Oishi, T., Nakazawa, A., Kurazume, R., Ikeuchi, K., & Sagawa, R. (2008). Parallel alignment of a large number of range images. In Digitally Archiving Cultural Objects (pp. 109-126). Springer US. https://doi.org/10.1007/978-0-387-75807_7

Parallel alignment of a large number of range images. / Oishi, Takeshi; Nakazawa, Atsushi; Kurazume, Ryo; Ikeuchi, Katsushi; Sagawa, Ryusuke.

Digitally Archiving Cultural Objects. Springer US, 2008. p. 109-126.

Research output: Chapter in Book/Report/Conference proceedingChapter

Oishi, T, Nakazawa, A, Kurazume, R, Ikeuchi, K & Sagawa, R 2008, Parallel alignment of a large number of range images. in Digitally Archiving Cultural Objects. Springer US, pp. 109-126. https://doi.org/10.1007/978-0-387-75807_7
Oishi T, Nakazawa A, Kurazume R, Ikeuchi K, Sagawa R. Parallel alignment of a large number of range images. In Digitally Archiving Cultural Objects. Springer US. 2008. p. 109-126 https://doi.org/10.1007/978-0-387-75807_7
Oishi, Takeshi ; Nakazawa, Atsushi ; Kurazume, Ryo ; Ikeuchi, Katsushi ; Sagawa, Ryusuke. / Parallel alignment of a large number of range images. Digitally Archiving Cultural Objects. Springer US, 2008. pp. 109-126
@inbook{f62357360c1649d8b521d0e13bdb0dc7,
title = "Parallel alignment of a large number of range images",
abstract = "This chapter describes a method for parallel alignment of multiple range images. There are problems of computational time and memory space in aligning a large number of range images simultaneously. We developed a parallel method to address the problems. Searching for corresponding points between two range images is time-consuming and requires considerable memory space when performed independently. However, this process can be preformed in parallel, with each corresponding pair of range images assigned to a node. Because the computation time is approximately proportional to the number of vertices, by assigning the pairs so that the number of vertices computed is equal on each node, the load on each node is effectively distributed. In order to reduce the amount of memory required on each node, a hypergraph that represents the correspondences of range images is created, and heuristic graph partitioning algorithms are applied to determine the optimal assignment of the pairs. Moreover, by rejecting redundant dependencies, it becomes possible to accelerate computation time and reduce the amount of memory required on each node. The method was tested on a 16-processor PC cluster, where it demonstrated high extendibility and improved performance.",
author = "Takeshi Oishi and Atsushi Nakazawa and Ryo Kurazume and Katsushi Ikeuchi and Ryusuke Sagawa",
year = "2008",
month = "12",
day = "1",
doi = "10.1007/978-0-387-75807_7",
language = "English",
isbn = "9780387758060",
pages = "109--126",
booktitle = "Digitally Archiving Cultural Objects",
publisher = "Springer US",
address = "United States",

}

TY - CHAP

T1 - Parallel alignment of a large number of range images

AU - Oishi, Takeshi

AU - Nakazawa, Atsushi

AU - Kurazume, Ryo

AU - Ikeuchi, Katsushi

AU - Sagawa, Ryusuke

PY - 2008/12/1

Y1 - 2008/12/1

N2 - This chapter describes a method for parallel alignment of multiple range images. There are problems of computational time and memory space in aligning a large number of range images simultaneously. We developed a parallel method to address the problems. Searching for corresponding points between two range images is time-consuming and requires considerable memory space when performed independently. However, this process can be preformed in parallel, with each corresponding pair of range images assigned to a node. Because the computation time is approximately proportional to the number of vertices, by assigning the pairs so that the number of vertices computed is equal on each node, the load on each node is effectively distributed. In order to reduce the amount of memory required on each node, a hypergraph that represents the correspondences of range images is created, and heuristic graph partitioning algorithms are applied to determine the optimal assignment of the pairs. Moreover, by rejecting redundant dependencies, it becomes possible to accelerate computation time and reduce the amount of memory required on each node. The method was tested on a 16-processor PC cluster, where it demonstrated high extendibility and improved performance.

AB - This chapter describes a method for parallel alignment of multiple range images. There are problems of computational time and memory space in aligning a large number of range images simultaneously. We developed a parallel method to address the problems. Searching for corresponding points between two range images is time-consuming and requires considerable memory space when performed independently. However, this process can be preformed in parallel, with each corresponding pair of range images assigned to a node. Because the computation time is approximately proportional to the number of vertices, by assigning the pairs so that the number of vertices computed is equal on each node, the load on each node is effectively distributed. In order to reduce the amount of memory required on each node, a hypergraph that represents the correspondences of range images is created, and heuristic graph partitioning algorithms are applied to determine the optimal assignment of the pairs. Moreover, by rejecting redundant dependencies, it becomes possible to accelerate computation time and reduce the amount of memory required on each node. The method was tested on a 16-processor PC cluster, where it demonstrated high extendibility and improved performance.

UR - http://www.scopus.com/inward/record.url?scp=84891462279&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84891462279&partnerID=8YFLogxK

U2 - 10.1007/978-0-387-75807_7

DO - 10.1007/978-0-387-75807_7

M3 - Chapter

SN - 9780387758060

SP - 109

EP - 126

BT - Digitally Archiving Cultural Objects

PB - Springer US

ER -