Efficient arctangent computation for real-time histograms of oriented gradients descriptor extraction

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

In this paper, we propose the hardware architecture for the implementation of arctangent operation in Histograms of Oriented Gradients (HOG) descriptor extraction, targeting real-time pedestrian recognition with high accuracy on an embedded system with feasible hardware complexity.

Original languageEnglish
Title of host publicationComputational Intelligence and Intelligent Systems - 6th International Symposium, ISICA 2012, Proceedings
Pages612-616
Number of pages5
DOIs
StatePublished - 2012
Event6th International Symposium on Intelligence Computation and Applications, ISICA 2012 - Wuhan, China
Duration: 27 Oct 201228 Oct 2012

Publication series

NameCommunications in Computer and Information Science
Volume316 CCIS
ISSN (Print)1865-0929

Conference

Conference6th International Symposium on Intelligence Computation and Applications, ISICA 2012
Country/TerritoryChina
CityWuhan
Period27/10/1228/10/12

Keywords

  • Arctangent Operation
  • Hardware Accelerator
  • Histograms of Oriented Gradients (HOG)
  • Pedestrian Recognition
  • System-on-Chip (SoC)

Fingerprint

Dive into the research topics of 'Efficient arctangent computation for real-time histograms of oriented gradients descriptor extraction'. Together they form a unique fingerprint.

Cite this