Hexa Cam: An FPGA-Based Multi-view Camera System

Abraham Monrroy, Manato Hirabayashi, Shinpei Kato, Masato Edahiro, Takefumi Miyoshi, Satoshi Funada

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Accurate perception information of the surroundings is a desirable feature in any cyber-physical system for autonomous robotics. Besides, providing a cost-wise solution to this need, is also useful for developing other kinds of applications such as mapping, decision-making, self-localization, among others. In this work, we present and benchmark a cost-effective camera device with six image sensors attached to a custom built FPGA board. The perception information is delivered on the host side, thanks to a HOG based object detector executed on the GPU. Our conclusion is that the low cost prototype is able to acquire and extract the perception data in 9.41 frames per second.

Original languageEnglish
Title of host publicationProceedings - 3rd IEEE International Conference on Cyber-Physical Systems, Networks, and Applications, CPSNA 2015
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages48-53
Number of pages6
ISBN (Electronic)9781467377850
DOIs
Publication statusPublished - 2015 Sep 18
Externally publishedYes
Event3rd IEEE International Conference on Cyber-Physical Systems, Networks, and Applications, CPSNA 2015 - Hong Kong, China
Duration: 2015 Aug 192015 Aug 21

Publication series

NameProceedings - 3rd IEEE International Conference on Cyber-Physical Systems, Networks, and Applications, CPSNA 2015

Conference

Conference3rd IEEE International Conference on Cyber-Physical Systems, Networks, and Applications, CPSNA 2015
CountryChina
CityHong Kong
Period15/8/1915/8/21

Keywords

  • Camera sensor
  • FPGA
  • GPU
  • Object Detection
  • Perception
  • Real-Time

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Hardware and Architecture

Fingerprint Dive into the research topics of 'Hexa Cam: An FPGA-Based Multi-view Camera System'. Together they form a unique fingerprint.

Cite this