Work

Online LiDAR-Camera Extrinsic Calibration Using Selected Semantic Features

Year: 2025

Type: article

Abstract: Autonomous vehicles have gained great attention from all walks of life in recent years. The relative position and orientation between sensors often change gradually over time due to vibrations or thermal stress of materials. Thus, online re-calibrating extrinsic parameters periodically is required. In this situation, automatic targetless methods are more preferable as they do not require a calibration target or tedious calibration procedure. In this paper, we propose an online targetless camera-LiDAR extrinsic calibration approach with the help of semantic information. Our method could effectively ameliorate the problem of targetless methods which usually lack robust features and the correspondences. We also propose a feature selection technique to filter out improper feature correspondences by matching the image contours and point cloud projection contours. The experiment results show that our approach is more robust than previous work, and the calibration algorithm is applicable to more scenarios.

Source: IEEE Open Journal of Intelligent Transportation Systems

DOI: 10.1109/ojits.2025.3555574

Authors: Ping-Tzu Lin, Ying-Shiuan Huang, Wen-Chieh Lin, Chieh-Chih Wang, Huei-Yung Lin

Institutions: National Yang Ming Chiao Tung University, National Taipei University of Technology

Language: English

Topic: Robotics and Sensor-Based Localization

Subfield: Aerospace Engineering

Field: Engineering

Domain: Physical Sciences

Open Access status: gold

Estimated APC paid: $2,075

Funder: National Science and Technology Council