Skip to main navigation Skip to search Skip to main content

Road Detection Based on Region Grid Analysis Using Structural Similarity

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

Abstract

Road detection is one of common used in autonomous vehicle robot. The fast road detection process is needed by autonomous vehicle robot for steering the robot following the road. So, in this paper proposes a method to detect roads using similarity analysis. An image is obtained by capturing from the camera sensor. This image is graded and then the region of image in the grid areas is analyzed by structural similarity (SSIM) index analysis. The structural similarity method compares the two images between region grid image area and references image. A reference image is a sample image that represents a road surface. In this paper, we accept the region grid image areas as a road surface, if it has SSIM index value greater than 0.995. This value is obtained from the results of repeated experiments. The experiment shows that the proposed method has capable detect the road surface in the region of the grid image areas, as shown in the image in the experimental result.

Original languageEnglish
Title of host publication2020 IEEE 4th International Conference on Frontiers of Sensors Technologies, ICFST 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages63-66
Number of pages4
ISBN (Electronic)9781728186009
DOIs
Publication statusPublished - 6 Nov 2020
Event4th IEEE International Conference on Frontiers of Sensors Technologies, ICFST 2020 - Virtual, Shanghai, China
Duration: 6 Nov 20209 Nov 2020

Publication series

Name2020 IEEE 4th International Conference on Frontiers of Sensors Technologies, ICFST 2020

Conference

Conference4th IEEE International Conference on Frontiers of Sensors Technologies, ICFST 2020
Country/TerritoryChina
CityVirtual, Shanghai
Period6/11/209/11/20

Keywords

  • image capturing
  • region grid
  • resizing
  • SSIM

Fingerprint

Dive into the research topics of 'Road Detection Based on Region Grid Analysis Using Structural Similarity'. Together they form a unique fingerprint.

Cite this