ROI-based Calibration Algorithm of Camera-Lidar Sensor Fusion That is Strong in Object Recognition and Distance Information Extraction

Young Jin Yoon, Dong Seog Han

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

In an autonomous vehicle, accurate location recognition with an object on the road is a very important factor for driving safety. In this paper, we propose an region of interest (ROI)-based location information extraction algorithm that converges a camera and a lidar for accurate location extraction of an object. In general, when the camera and lidar are fused to match the position of an object, a target board such as a chess board is used to calibrate the position difference between the sensors. In the area of interest for object recognition of an autonomous vehicle, a difference in the positions of the camera and lidar occurs depending on the position of the object due to the wide road environment. The proposed algorithm divides the camera image into multiple ROIs and performs the position matching of the camera and lidar independently for each area. We verify the accuracy of the proposed ROI-based positioning algorithm in virtual and real road environment.

Original languageEnglish
Pages (from-to)2301-2309
Number of pages9
JournalJournal of Korean Institute of Communications and Information Sciences
Volume46
Issue number12
DOIs
StatePublished - Dec 2021

Keywords

  • Autonomous driving
  • Calibration
  • Object detection
  • Sensor Fusion

Fingerprint

Dive into the research topics of 'ROI-based Calibration Algorithm of Camera-Lidar Sensor Fusion That is Strong in Object Recognition and Distance Information Extraction'. Together they form a unique fingerprint.

Cite this