Observatorio de I+D+i UPM

Memorias de investigación
Communications at congresses:
Integration of PCA and SVM on a Massively Parallel Processor Array (MPPA) platform applied to hyperspectral images
Year:2016
Research Areas
  • Electric engineers, electronic and automatic (eil)
Information
Abstract
In this paper, a study of the parallel adaptation of a Principal Component Analysis (PCA) algorithm and a S upport Vector Machine (S VM) classifier, both running on a Massively Parallel Processor Array (MPPA) platform ?a system that assembles 256 cores in 16 different clusters?, is presented. The aim of this research has been to exploit the potential offered by a manycore MPPA platform to implement in parallel both PCA and S VM, so as to minimize the required time to process a hyperspectral image. These two algorithms are a key part of a more general one intended to discriminate between cancer and normal tissues during neurosurgical procedures. Experimenting with medical brain images captured in two operating theaters, the processing time measured when executing both algorithms simultaneously has been compared to the one obtained when executing the same algorithms sequentially. As a result, an average speedup of 6.19 has been achieved. Consequently, these algorithms consume less than an 8% of the available time to process a hyperspectral image, considering this time as the one required for the hyperspectral sensor to capture a new image.
International
Si
Congress
XXXI Design of Circuits and Integrated Systems Conference
960
Place
Granada, Spain
Reviewers
Si
ISBN/ISSN
978-1-5090-4565-5
Start Date
23/11/2016
End Date
25/11/2016
From page
1
To page
6
Proceedings of the XXXI Design of Circuits and Integrated Systems Conference
Participants
  • Autor: Daniel Madroñal Quintin (UPM)
  • Autor: Raquel Lazcano Lopez (UPM)
  • Autor: S. Ortega (IUMA-ULPGC)
  • Autor: H. Fabelo (IUMA-ULPGC)
  • Autor: Ruben Salvador Perea (UPM)
  • Autor: Eduardo Juarez Martinez (UPM)
  • Autor: G. M. Callicó (IUMA-ULPGC)
  • Autor: Cesar Sanz Alvaro (UPM)
Research Group, Departaments and Institutes related
  • Creador: Grupo de Investigación: Grupo de Diseño Electrónico y Microelectrónico
  • Departamento: Ingeniería Telemática y Electrónica
  • Centro o Instituto I+D+i: Tecnologías del Software y Sistemas Multimedia para la Sostenibilidad (CITSEM)
S2i 2020 Observatorio de investigación @ UPM con la colaboración del Consejo Social UPM
Cofinanciación del MINECO en el marco del Programa INNCIDE 2011 (OTR-2011-0236)
Cofinanciación del MINECO en el marco del Programa INNPACTO (IPT-020000-2010-22)