Application-specific instruction processor for extracting local binary patterns

Jani Boutellier; Ismo Lundbom; Janne Janhunen; Jori Ylimainen; Jari Hannuksela

Application-specific instruction processor for extracting local binary patterns

Jani Boutellier^*
, Ismo Lundbom
, Janne Janhunen
, Jori Ylimainen
, Jari Hannuksela

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

9 Citations (Scopus)

Abstract

Local Binary Pattern (LBP) is texture operator used in preprocessing for object detection, tracking, face recognition and fingerprint matching. Many of these applications are performed on embedded devices, which poses limitations on the implementation complexity and power consumption. As LBP features are computed pixelwise, high performance is required for real time extraction of LBP features from high resolution video. This paper presents an application-specific instruction processor for LBP extraction. The compact, yet powerful processor is capable of extracting LBP features from 1280 × 720p (30 fps) video with a reasonable 304 MHz clock rate. With a low power consumption and an area of less than 16k gates the processor is suitable for embedded devices. Experiments present resource and power consumption measured on an FPGA board, along with processor synthesis results. In terms of latency, our processor requires 17.5 × less clock cycles per LBP feature than a workstation implementation and only 2.0 × more than a hardwired ASIC.

Original language	English
Title of host publication	DASIP 2012 - Proceedings of the 2012 Conference on Design and Architectures for Signal and Image Processing
Pages	82-89
Number of pages	8
Publication status	Published - 2012
Publication type	A4 Article in conference proceedings
Event	6th Annual Conference on Design and Architectures for Signal and Image Processing, DASIP 2012 - Karlsruhe, Germany Duration: 23 Oct 2012 → 25 Oct 2012

Conference

Conference	6th Annual Conference on Design and Architectures for Signal and Image Processing, DASIP 2012
Country/Territory	Germany
City	Karlsruhe
Period	23/10/12 → 25/10/12

Keywords

Digital signal processors
Feature extraction
Image texture analysis
Video signal processing

ASJC Scopus subject areas

Computer Graphics and Computer-Aided Design
Computer Vision and Pattern Recognition
Hardware and Architecture
Signal Processing
Electrical and Electronic Engineering

Cite this

@inproceedings{64235ec3b02546509c5eec6182306bfe,

title = "Application-specific instruction processor for extracting local binary patterns",

abstract = "Local Binary Pattern (LBP) is texture operator used in preprocessing for object detection, tracking, face recognition and fingerprint matching. Many of these applications are performed on embedded devices, which poses limitations on the implementation complexity and power consumption. As LBP features are computed pixelwise, high performance is required for real time extraction of LBP features from high resolution video. This paper presents an application-specific instruction processor for LBP extraction. The compact, yet powerful processor is capable of extracting LBP features from 1280 × 720p (30 fps) video with a reasonable 304 MHz clock rate. With a low power consumption and an area of less than 16k gates the processor is suitable for embedded devices. Experiments present resource and power consumption measured on an FPGA board, along with processor synthesis results. In terms of latency, our processor requires 17.5 × less clock cycles per LBP feature than a workstation implementation and only 2.0 × more than a hardwired ASIC.",

keywords = "Digital signal processors, Feature extraction, Image texture analysis, Video signal processing",

author = "Jani Boutellier and Ismo Lundbom and Janne Janhunen and Jori Ylimainen and Jari Hannuksela",

year = "2012",

language = "English",

isbn = "9782953998726",

pages = "82--89",

booktitle = "DASIP 2012 - Proceedings of the 2012 Conference on Design and Architectures for Signal and Image Processing",

note = "6th Annual Conference on Design and Architectures for Signal and Image Processing, DASIP 2012 ; Conference date: 23-10-2012 Through 25-10-2012",

}

Boutellier, J, Lundbom, I, Janhunen, J, Ylimainen, J & Hannuksela, J 2012, Application-specific instruction processor for extracting local binary patterns. in DASIP 2012 - Proceedings of the 2012 Conference on Design and Architectures for Signal and Image Processing., 6385363, pp. 82-89, 6th Annual Conference on Design and Architectures for Signal and Image Processing, DASIP 2012, Karlsruhe, Germany, 23/10/12.

Application-specific instruction processor for extracting local binary patterns. / Boutellier, Jani; Lundbom, Ismo; Janhunen, Janne et al.
DASIP 2012 - Proceedings of the 2012 Conference on Design and Architectures for Signal and Image Processing. 2012. p. 82-89 6385363.

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Application-specific instruction processor for extracting local binary patterns

AU - Boutellier, Jani

AU - Lundbom, Ismo

AU - Janhunen, Janne

AU - Ylimainen, Jori

AU - Hannuksela, Jari

PY - 2012

Y1 - 2012

N2 - Local Binary Pattern (LBP) is texture operator used in preprocessing for object detection, tracking, face recognition and fingerprint matching. Many of these applications are performed on embedded devices, which poses limitations on the implementation complexity and power consumption. As LBP features are computed pixelwise, high performance is required for real time extraction of LBP features from high resolution video. This paper presents an application-specific instruction processor for LBP extraction. The compact, yet powerful processor is capable of extracting LBP features from 1280 × 720p (30 fps) video with a reasonable 304 MHz clock rate. With a low power consumption and an area of less than 16k gates the processor is suitable for embedded devices. Experiments present resource and power consumption measured on an FPGA board, along with processor synthesis results. In terms of latency, our processor requires 17.5 × less clock cycles per LBP feature than a workstation implementation and only 2.0 × more than a hardwired ASIC.

AB - Local Binary Pattern (LBP) is texture operator used in preprocessing for object detection, tracking, face recognition and fingerprint matching. Many of these applications are performed on embedded devices, which poses limitations on the implementation complexity and power consumption. As LBP features are computed pixelwise, high performance is required for real time extraction of LBP features from high resolution video. This paper presents an application-specific instruction processor for LBP extraction. The compact, yet powerful processor is capable of extracting LBP features from 1280 × 720p (30 fps) video with a reasonable 304 MHz clock rate. With a low power consumption and an area of less than 16k gates the processor is suitable for embedded devices. Experiments present resource and power consumption measured on an FPGA board, along with processor synthesis results. In terms of latency, our processor requires 17.5 × less clock cycles per LBP feature than a workstation implementation and only 2.0 × more than a hardwired ASIC.

KW - Digital signal processors

KW - Feature extraction

KW - Image texture analysis

KW - Video signal processing

UR - https://www.scopus.com/pages/publications/84872397244

M3 - Conference contribution

AN - SCOPUS:84872397244

SN - 9782953998726

SP - 82

EP - 89

BT - DASIP 2012 - Proceedings of the 2012 Conference on Design and Architectures for Signal and Image Processing

T2 - 6th Annual Conference on Design and Architectures for Signal and Image Processing, DASIP 2012

Y2 - 23 October 2012 through 25 October 2012

ER -

Application-specific instruction processor for extracting local binary patterns

Abstract

Conference

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this