Method for identifying cylindrical surface from multi-line laser radar point cloud data

Abstract

The invention relates to a method for identifying a cylindrical surface from multi-line laser radar point cloud data. The method comprises the steps that firstly a cylindrical surface fitting mathematical model is set; then, a cylindrical surface fitting pretreatment step is carried out, and a RANSAC algorithm and a rotation matrix are utilized to determine two-dimensional plane ellipse fitting coordinates of each line of laser beam section; a cylindrical surface fitting step is carried out, a least square method and a Lagrangian multiplier algorithm are utilized to construct and solve an ellipse optimization target equation, fitting points are randomly selected, and an optimal fitting ellipse is obtained through solving by utilizing a least square short axis inclination angle constraint condition; and finally, seven parameters of the cylindrical surface are obtained through singular value decomposition and optimization calculation. According to the invention, the problem of cylindrical surface identification encountered when the cylindrical surface is taken as a calibration object by the laser radar in the process of combined external calibration of the multi-line laser radar andthe camera in the field of environmental perception can be solved, and the cylindrical surface identification precision of the low-resolution multi-line laser radar is relatively high.

Description

technical field [0001] The invention belongs to the field of environment perception, and in particular relates to a method for identifying a cylindrical surface from multi-line laser radar point cloud data. Background technique [0002] Due to the limitations of purchase cost and use conditions, the early lidar has not received widespread attention and application. However, in recent years, with the rapid development of lidar technology, its application fields have been continuously expanded, especially in the field of environmental perception. The fusion of lidar and camera to perceive the external environment has become the mainstream method used by unmanned systems. The heterogeneous fusion of lidar and camera first needs to use calibration objects for joint external calibration, and the selection of calibration objects directly affects the fusion effect. The most common calibration object in the existing joint external calibration literature is a black and white checker...

AI Technical Summary

Problems solved by technology

In the real environment, it is impossible to effectively find objects similar to checkerboard features, but there are many objects with cylindrical features in the real environment, such as cylinders in buildings, bridge piers, water lighthouses, and cylindrical buoys, etc. object

Using the cylindrical surface in the real environment as the calibration object in the joint external calibration is rarely mentioned, mainly because the laser radar obtains sparse and incomplete cylindrical surface point cloud data. The existing research literature on cylindrical surface fitting is based on The high-precision and high-resolution 3D laser scanner is based on the roughly complete point cloud data of the cylindrical surface, and there is little research on the fitting of the sparse and incomplete point cloud cylindrical surface

Images