Multi-sensor Calibration Method For Mobile Robot

With the increasing application of mobile robots in industry,service industry,and medical industry,multi-sensor joint calibration,as one of the key technologies,has attracted much attention of researchers.Multi-sensor calibration realizes the unification of multi-sensor coordinate systems.It provides the basis for the subsequent work of mobile robots.This article focuses on lidar and camera,which are the most used sensors.This paper mainly researches from two aspects,one is refinement of edge extraction based on lidar point cloud,and the other is lidar-camera calibration and data fusion based on multi-angle.The specific tasks are as follows:(1)First,the research status and main methods of calibration of lidar and camera are introduced in detail.And the key technologies are discussed in depth.(2)About the refinement of edge extraction based on lidar point cloud,this paper mainly introduces the central idea and specific process of edge extraction.For the shortcomings of the direct method,the scanning characteristics of lidar are used to provide a refined edge point calculation strategy.Multiple data frames are then used to calculate points closer to the edge.Experiments show that the method proposed compensates for hardware deficiencies through algorithms and improves the accuracy of lidar edge points.(3)About the lidar-camera calibration and data fusion based on multi-angle,this paper introduces the calibration method based on Ar Uco and the data fusion method on multiple angles.The purpose of calibration is to project lidar points onto pixels.The projection result of the parameter obtained on a single angle is accurate only on its own angle.Due to this defect,the paper proposes a data fusion method on multiple angles.The distances from the projection point to different areas of the image and the errors of projection parameters are calculated.Then the optimized projection point is obtained by weighted calculation.Experiments show that this method effectively reduces the impact of angle transformation on projection accuracy,and the results are more accurate.