A laser radar and camera fusion device and calibration method

Abstract

The invention discloses a laser radar and camera fusion device and a calibration method, comprising a vertical lifting platform and a calibration plate support; a working platform is arranged on the working guide rail of the vertical lifting platform, and a camera is fixedly installed on the working platform and laser radar; the calibration plate bracket is fixedly installed with a calibration plate; the laser radar scans a horizontal straight line on the calibration plate, finds the left and right endpoints of the straight line on the boundary of the calibration plate, and then moves the same interval multiple times , you can get multiple sets of endpoints on the left and right borders, and the fitting can approximate the straight line equations on the left and right borders, and then get the endpoints of the calibration board; match the pixel points of the endpoints on the camera image with the endpoints to get the camera and radar. Pose relationship expression; using this laser radar and camera fusion calibration method can obtain the features of the scene where the radar point cloud is sparse, and can solve the joint expression of the internal and external parameters, without the need to measure the internal parameters of the camera in advance.

Description

technical field [0001] The invention relates to the technical field of fusion and calibration, in particular to a laser radar and camera fusion device and a calibration method. Background technique [0002] With the rise of computer technology, 3D reconstruction has gradually become a hot topic. Commonly used 3D reconstruction sensors are usually binocular cameras or lidars to complete object reconstruction. For the binocular camera, although it has color information, it will bring computational complexity to the entire system due to its complex matching between images, and only sparse point cloud 3D reconstruction can be performed. For lidar, the results are accurate and reliable, but because it does not acquire color information, the results of 3D reconstruction cannot be intuitively felt. [0003] Therefore, a single sensor inevitably has limitations. In order to improve the robustness of the system and fuse color information while ensuring accuracy, a multi-sensor fusio...

AI Technical Summary

Problems solved by technology

For the binocular camera, although it has color information, due to its complex image matching, it will bring computational complexity to the entire system and can only perform sparse point cloud 3D reconstruction

For lidar, its results are accurate and reliable, but because it does not acquire color information, it cannot be intuitively felt for the results of 3D reconstruction

Images