Extrinsic calibration of a laser range finder and a camera based on the automatic detection of line feature

This paper presents an extrinsic calibration method of a camera and a laser range finder based on the automatic detection of the line feature. Firstly, a special 3D calibration target which contained four planes and five edges was projected, and then the visible image and laser data of the 3D calibration target could be obtained. Secondly, the automatic detection method of the line feature was presented, which used the linear fitting to acquire the equations of the line features detected by the Hough transform. With the obtained equations, the laser data corresponding to the line features of the 3D calibration target could be acquired by computing the crossover point of two straight-lines. Finally, the rotation matrix and translation matrix between the camera and laser range finder could be acquired with computing the linear equations provided by the laser data and the corresponding crossover point. Experimental result showed the proposed method could acquire the accurate external calibrations.

I. Introduction

The laser range finder has increasingly been used in machine vision since it could provide the depth information of the object or the scene, however the laser range finder can't acquire the texture. To acquire the intact 3D reconstruction result of the object or the scene, the extrinsic calibration of a camera and a laser range finder has been the research focus of the machine vision, image processing and pattern recognition, which has been widely application in mobile robot, aeronautics and astronautics, historic preservation^[1]–[4].