Automatic Calibration of Camera and LRF based on Morphological Pattern and Optimal Angular Back-projection Error

Nowadays, integration systems based on cameras and LRFs have been applied in many researches on robotics, such as autonomous navigation vehicles, and intelligent transportation systems. The systems using multiple sensors usually require extracting the-relative information of sensors posed, which support for processing. Therefore, the task of calibration of cameras and laser devices is very important. This paper presents an automatic calibration method for determining the relative position and direction of a camera with respect to a laser rangefinder. The calibration method makes use of depth discontinuities of calibration patterns, which emphasizes laser range-beams supporting easy automatic estimation of the occurred position of laser scans on the calibration pattern. The laser range scans are also used for estimating corresponding depth image points from the camera. Finally, the relative parameters between the camera and the laser device are discovered using the set of depth point correspondences of the LRF and the camera. This paper also presents new criterion for evaluating the error of extrinsic parameters estimation based on optimal error of angular deviation between the projection and back-projection of 3D corresponding points. The evaluation results demonstrate outperformance of this method.