Targetless Camera-LiDAR Calibration in Unstructured Environments

Abstract

The camera-Lidar sensor fusion plays an important role in autonomous navigation research. Nowadays, the automatic calibration of these sensors remains a significant challenge in mobile robotics. In this article, we present a novel calibration method that achieves an accurate six-degree-offreedom (6-DOF) rigid-body transformation estimation (aka extrinsic parameters) between the camera and LiDAR sensors. This method consists of a novel co-registration approach that uses local edge features in arbitrary environments to get 3D-to-2D errors between the data of both, camera and LiDAR. Once we have 3D-to-2D errors, we estimate the relative transform, i.e., the extrinsic parameters, that minimizes these errors. In order to find the best transform solution, we use the perspective-three-point (P3P) algorithm. To refine the final calibration, we use a Kalman Filter, which gives the system high stability against noise disturbances. The presented method does not require, in any case, an artificial target, or a structured environment, and therefore, it is a target-less calibration. Furthermore, the method we present in this article does not require to achieve a dense point cloud, which holds the advantage of not needing a scan accumulation. To test our approach, we use the state-of-the-art Kitti dataset, taking the calibration provided by the dataset as the ground truth. In this way, we achieve accuracy results, and we demonstrate the robustness of the system against very noisy observations.

Automatic Multiple LiDAR Calibration Based on the Plane Features of Structured Environments

“…To solve the problem mentioned in the previous subsection, many algorithms were proposed to perform automatic online calibration by extracting the features of surrounding environments [28][29][30][31][32]. Pandey et al [28] proposed a mutual information-based calibration method using the correlation between laser reflectivity and camera intensity.…”

“…Pandey et al [28] proposed a mutual information-based calibration method using the correlation between laser reflectivity and camera intensity. Also, Levinson et al [29] and Muñoz-Bañón et al [30] proposed an automatic calibration method using the edge information extracted from each camera and LiDAR sensor. For multi-LiDAR calibrations, Gao et al [31] suggested a calibration method by limiting feature information using the reflectance information of LiDAR in the retro reflection materials attached to poles.…”

DA-LMR: A Robust Lane Marking Representation for Data Association

“…Distance-Compatible Sample Consensus (DC-SAC) could be seen as a heuristic version of the RANSAC method. This is a formalization of a previously used in [4] for camera-LiDAR calibration approach. In [38], the authors propose a similar approach but focus on dense point clouds and geometric descriptors.…”

“…• DA-LMR, a new way of representing lane markings for data association. • A formalization of the data association method DC-SAC, used in our previous work [4]. • An evaluation of the possible combinations of data representation and data association common state-ofthe-art approaches, including those proposed in this paper.…”

TEScalib: Targetless Extrinsic Self-Calibration of LiDAR and Stereo Camera for Automated Driving Vehicles with Uncertainty Analysis

“…In the work [1], the authors propose a targetless LiDAR and camera extrinsic calibration approach based on mutual information (MI) using LiDAR intensity measurements and ogy, Karlsruhe, Germany. {haohao.hu, janhendrik.pauls, stiller}@kit.edu, skyhfz@gmail.com 2 Mobile Perception Systems Department, FZI Research Center for Information Technology, Karlsruhe, Germany. bieder@fzi.de Fig.…”