CUDA-Based Computation for Visual Odometry

Shen Ho Liu, Chen Chien Hsu, Wei Yen Wang, Cheng Hung Lin

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

Abstract

An enhanced visual odometry (VO) system is proposed to improve the accuracy of pose estimation based on a corrected model, and the matching algorithm is implemented on graphical processing units (GPUs) so that the computation can be accelerated in parallel and in real-time using the compute unified device architecture (CUDA) programming model. To evaluate the performance of the proposed approach, an ASUS Xtion 3D camera, laptop, and NVIDIA TX2 are employed to conduct extensive experiments. The experimental results show that compared with the traditional VO algorithm, the proposed approach gives better results over the traditional VO algorithm.

Original languageEnglish
Title of host publication2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages55-56
Number of pages2
ISBN (Electronic)9781538663097
DOIs
Publication statusPublished - 2018 Dec 12
Event7th IEEE Global Conference on Consumer Electronics, GCCE 2018 - Nara, Japan
Duration: 2018 Oct 92018 Oct 12

Publication series

Name2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018

Other

Other7th IEEE Global Conference on Consumer Electronics, GCCE 2018
CountryJapan
CityNara
Period18/10/918/10/12

    Fingerprint

Keywords

  • CUDA
  • GPU
  • Visual odometry

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Electrical and Electronic Engineering
  • Safety, Risk, Reliability and Quality
  • Instrumentation

Cite this

Liu, S. H., Hsu, C. C., Wang, W. Y., & Lin, C. H. (2018). CUDA-Based Computation for Visual Odometry. In 2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018 (pp. 55-56). [8574869] (2018 IEEE 7th Global Conference on Consumer Electronics, GCCE 2018). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/GCCE.2018.8574869