Accelerating multiresolution gabor feature extraction for real time vision applications

Yong Cheol Peter Cho, Nandhini Chandramoorthy, Kevin M. Irick, Vijaykrishnan Narayanan

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Multiresolution Gabor filter banks are used for feature extraction in a variety of applications as Gabor filters have shown to be exceptional feature extractors with a close correspondence to the simple cells in the primary visual cortex (V1) of the brain. Yet applying the Gabor filter is a computationally intensive task. Most applications that utilize the Gabor feature space require real time results; however, the large quantity of computations involved has hindered systems from achieving real time performance. The natural solution for such compute intensive tasks is parallelization. FPGAs have emerged as attractive platforms for compute intensive signal processing applications due to their massively parallel computation resources as well as low power consumption and affordability. We present a configurable architecture for Gabor feature extraction on FPGA that enhances the resource utilization of the FPGA hardware fabric while maintaining a streaming data flow to yield exceptional performance. The increased resource utilization resulting from configurability, optimizations, and resource sharing allows for higher levels of parallelism to achieve real time feature extraction of high resolution images. Two architectures are introduced. The first is an architecture for multiresolution feature extraction with extensive resource sharing for enhanced resource utilization. The second is an architecture for many-orientation applications using a coarse to fine grain method to enhance resource utilization by reducing the number of filters applied at different orientations. Our results show that our multiresolution implementation achieves real-time performance on 2048×1526 images and exhibits 6X speed up over a GPU implementation while exhibiting energy efficiency with 0.4fps/W compared to the GPU that achieves 0.036fps/W.[1] The implementation for many-orientation applications using the coarse to fine grain method exhibits resource saving of at most 2√O for O number of orientations and higher, compared to a fully parallel architecture and 25× speedup compared to a GPU implementation for 16 orientations.

Original languageEnglish (US)
Pages (from-to)149-168
Number of pages20
JournalJournal of Signal Processing Systems
Volume76
Issue number2
DOIs
StatePublished - Jan 1 2014

Fingerprint

Multiresolution
Feature Extraction
Feature extraction
Gabor filters
Gabor Filter
Resources
Field programmable gate arrays (FPGA)
Field Programmable Gate Array
Resource Sharing
Speedup
Parallel architectures
Filter banks
Streaming Data
Visual Cortex
Image resolution
Extractor
Filter Banks
Parallel Architectures
Parallel Computation
Energy efficiency

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Theoretical Computer Science
  • Signal Processing
  • Information Systems
  • Modeling and Simulation
  • Hardware and Architecture

Cite this

Cho, Yong Cheol Peter ; Chandramoorthy, Nandhini ; Irick, Kevin M. ; Narayanan, Vijaykrishnan. / Accelerating multiresolution gabor feature extraction for real time vision applications. In: Journal of Signal Processing Systems. 2014 ; Vol. 76, No. 2. pp. 149-168.
@article{4edc4ee58c984bb6a35c7f4bd95c4b02,
title = "Accelerating multiresolution gabor feature extraction for real time vision applications",
abstract = "Multiresolution Gabor filter banks are used for feature extraction in a variety of applications as Gabor filters have shown to be exceptional feature extractors with a close correspondence to the simple cells in the primary visual cortex (V1) of the brain. Yet applying the Gabor filter is a computationally intensive task. Most applications that utilize the Gabor feature space require real time results; however, the large quantity of computations involved has hindered systems from achieving real time performance. The natural solution for such compute intensive tasks is parallelization. FPGAs have emerged as attractive platforms for compute intensive signal processing applications due to their massively parallel computation resources as well as low power consumption and affordability. We present a configurable architecture for Gabor feature extraction on FPGA that enhances the resource utilization of the FPGA hardware fabric while maintaining a streaming data flow to yield exceptional performance. The increased resource utilization resulting from configurability, optimizations, and resource sharing allows for higher levels of parallelism to achieve real time feature extraction of high resolution images. Two architectures are introduced. The first is an architecture for multiresolution feature extraction with extensive resource sharing for enhanced resource utilization. The second is an architecture for many-orientation applications using a coarse to fine grain method to enhance resource utilization by reducing the number of filters applied at different orientations. Our results show that our multiresolution implementation achieves real-time performance on 2048×1526 images and exhibits 6X speed up over a GPU implementation while exhibiting energy efficiency with 0.4fps/W compared to the GPU that achieves 0.036fps/W.[1] The implementation for many-orientation applications using the coarse to fine grain method exhibits resource saving of at most 2√O for O number of orientations and higher, compared to a fully parallel architecture and 25× speedup compared to a GPU implementation for 16 orientations.",
author = "Cho, {Yong Cheol Peter} and Nandhini Chandramoorthy and Irick, {Kevin M.} and Vijaykrishnan Narayanan",
year = "2014",
month = "1",
day = "1",
doi = "10.1007/s11265-014-0873-4",
language = "English (US)",
volume = "76",
pages = "149--168",
journal = "Journal of Signal Processing Systems",
issn = "1939-8018",
publisher = "Springer New York",
number = "2",

}

Accelerating multiresolution gabor feature extraction for real time vision applications. / Cho, Yong Cheol Peter; Chandramoorthy, Nandhini; Irick, Kevin M.; Narayanan, Vijaykrishnan.

In: Journal of Signal Processing Systems, Vol. 76, No. 2, 01.01.2014, p. 149-168.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Accelerating multiresolution gabor feature extraction for real time vision applications

AU - Cho, Yong Cheol Peter

AU - Chandramoorthy, Nandhini

AU - Irick, Kevin M.

AU - Narayanan, Vijaykrishnan

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Multiresolution Gabor filter banks are used for feature extraction in a variety of applications as Gabor filters have shown to be exceptional feature extractors with a close correspondence to the simple cells in the primary visual cortex (V1) of the brain. Yet applying the Gabor filter is a computationally intensive task. Most applications that utilize the Gabor feature space require real time results; however, the large quantity of computations involved has hindered systems from achieving real time performance. The natural solution for such compute intensive tasks is parallelization. FPGAs have emerged as attractive platforms for compute intensive signal processing applications due to their massively parallel computation resources as well as low power consumption and affordability. We present a configurable architecture for Gabor feature extraction on FPGA that enhances the resource utilization of the FPGA hardware fabric while maintaining a streaming data flow to yield exceptional performance. The increased resource utilization resulting from configurability, optimizations, and resource sharing allows for higher levels of parallelism to achieve real time feature extraction of high resolution images. Two architectures are introduced. The first is an architecture for multiresolution feature extraction with extensive resource sharing for enhanced resource utilization. The second is an architecture for many-orientation applications using a coarse to fine grain method to enhance resource utilization by reducing the number of filters applied at different orientations. Our results show that our multiresolution implementation achieves real-time performance on 2048×1526 images and exhibits 6X speed up over a GPU implementation while exhibiting energy efficiency with 0.4fps/W compared to the GPU that achieves 0.036fps/W.[1] The implementation for many-orientation applications using the coarse to fine grain method exhibits resource saving of at most 2√O for O number of orientations and higher, compared to a fully parallel architecture and 25× speedup compared to a GPU implementation for 16 orientations.

AB - Multiresolution Gabor filter banks are used for feature extraction in a variety of applications as Gabor filters have shown to be exceptional feature extractors with a close correspondence to the simple cells in the primary visual cortex (V1) of the brain. Yet applying the Gabor filter is a computationally intensive task. Most applications that utilize the Gabor feature space require real time results; however, the large quantity of computations involved has hindered systems from achieving real time performance. The natural solution for such compute intensive tasks is parallelization. FPGAs have emerged as attractive platforms for compute intensive signal processing applications due to their massively parallel computation resources as well as low power consumption and affordability. We present a configurable architecture for Gabor feature extraction on FPGA that enhances the resource utilization of the FPGA hardware fabric while maintaining a streaming data flow to yield exceptional performance. The increased resource utilization resulting from configurability, optimizations, and resource sharing allows for higher levels of parallelism to achieve real time feature extraction of high resolution images. Two architectures are introduced. The first is an architecture for multiresolution feature extraction with extensive resource sharing for enhanced resource utilization. The second is an architecture for many-orientation applications using a coarse to fine grain method to enhance resource utilization by reducing the number of filters applied at different orientations. Our results show that our multiresolution implementation achieves real-time performance on 2048×1526 images and exhibits 6X speed up over a GPU implementation while exhibiting energy efficiency with 0.4fps/W compared to the GPU that achieves 0.036fps/W.[1] The implementation for many-orientation applications using the coarse to fine grain method exhibits resource saving of at most 2√O for O number of orientations and higher, compared to a fully parallel architecture and 25× speedup compared to a GPU implementation for 16 orientations.

UR - http://www.scopus.com/inward/record.url?scp=84907599600&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84907599600&partnerID=8YFLogxK

U2 - 10.1007/s11265-014-0873-4

DO - 10.1007/s11265-014-0873-4

M3 - Article

AN - SCOPUS:84907599600

VL - 76

SP - 149

EP - 168

JO - Journal of Signal Processing Systems

JF - Journal of Signal Processing Systems

SN - 1939-8018

IS - 2

ER -