Parallel image segmentation in reconfigurable chip multiprocessors

Authors:
Raphael Fonte Boa;Alexandre Marques Amaral;Dulcinéia Oliveira da Penha;Carlos Augusto P. da Silva Martins;Petr Y. Ekel
Affiliations:
Pontifical Catholic University of Minas Gerais (Brazil), MG, Brazil;Pontifical Catholic University of Minas Gerais (Brazil), MG, Brazil;Pontifical Catholic University of Minas Gerais (Brazil), MG, Brazil;Pontifical Catholic University of Minas Gerais (Brazil), MG, Brazil;Pontifical Catholic University of Minas Gerais (Brazil), MG, Brazil
Venue:
ISPA'06 Proceedings of the 2006 international conference on Frontiers of High Performance Computing and Networking
Year:
2006

Citing 7
Cited 0

Vision for Mobile Robot Navigation: A Survey

IEEE Transactions on Pattern Analysis and Machine Intelligence
Reconfigurable computing: a survey of systems and software

ACM Computing Surveys (CSUR)
A Single-Chip Multiprocessor

Computer
Efficient image processing applications on a network of workstations

CAMP '95 Proceedings of the Computer Architectures for Machine Perception
An FPGA Implementation of a Flexible, Parallel Image Processing Architecture Suitable for Embedded Vision Systems

IPDPS '03 Proceedings of the 17th International Symposium on Parallel and Distributed Processing
RCMP: a reconfigurable chip-multiprocessor architecture

ISPA'06 Proceedings of the 2006 international conference on Frontiers of High Performance Computing and Networking
Unsupervised segmentation of text fragments in real scenes

ICIAP'05 Proceedings of the 13th international conference on Image Analysis and Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Current image segmentation implementations are not optimized to all kinds of applications. To attend the different application kinds, the solution should allow to be reconfigured to fit their different characteristics and resource needs and to improve performance. Our objective is to present an image segmentation architecture and its implementation that can be reconfigured to execute different application workloads with demanded performance. In order to achieve this objective, our proposal is a parallel image segmentation implementation, which maps a pipelined parallel segmentation software architecture to a reconfigurable pipeline structure composed of reconfigurable chip multiprocessors (RCMPs). In this work, each pipeline stage was composed of a RCMP. Our results and its analysis show that our segmentation implementation provides greater flexibility and scalability and still obtains performance gain when compared to a multiprocessor machine. The main contribution is speedup, scalability and flexibility of the proposed solution.