Research On Indoor And Outdoor Point Cloud Registration Method Based On Point Cloud Orthophoto Feature Image

As the current main 3D spatial data information acquisition technology,3D laser scanning technology has been widely used in smart cities,3D printing,industrial production and other fields.When collecting spatial information,due to the limitation of the field of view of the measuring equipment,the shape of the measured object and the measurement environment,etc.,it is necessary to measure from multiple angles to obtain the complete three-dimensional spatial information of the measured object.In order to unify the three-dimensional spatial information of multiple perspectives into a unified coordinate system and complete the modeling,the registration of each point cloud is an indispensable and important link.The quality of the point cloud matching effect will directly affect the quality of the 3D model and its application.Existing algorithms have good registration accuracy when processing large-scale point cloud data registration,but their registration efficiency is low.In response to this problem,this paper uses the condition that the laser scanner is leveled during data acquisition or the built-in tilt sensor is used to compensate the tilt angle,and a four-degree-of-freedom point cloud initial registration framework based on point cloud orthographic feature images is proposed,and use the registration framework to conduct a registration study on challenging outdoor scene point clouds and indoor scene point clouds.The specific work content of this paper is as follows:(1)Introduced the relevant theories of point cloud registration technology.It mainly includes the mathematical model of the point cloud rigid body transformation,the method of solving the rigid body transformation parameters,and the four-degree-of-freedom point cloud registration,etc.The principle of the four-degree-of-freedom registration algorithm is mainly described,five commonly used four-degree-of-freedom algorithms are selected to introduce the principle,and the advantages and disadvantages of each method are analyzed.(2)Aiming at the problem that the existing algorithms cannot balance the registration accuracy and the registration efficiency when processing large-scale outdoor scene point cloud and indoor scene point cloud registration.This paper proposes a four-degree-of-freedom registration framework based on point cloud orthographic feature images,and uses this framework to propose an algorithm for outdoor point cloud registration and an algorithm for indoor point cloud registration.Five commonly used four-degree-of-freedom registration algorithms are used to conduct comparative experiments with the method proposed in this paper.Experimental results show that the two algorithms proposed in this paper are consistent with the traditional methods,and both can achieve centimeter-level registration accuracy.However,compared with the traditional algorithms,the registration efficiency of the method proposed in this paper is greatly improved.