Simplification Of 3D Laser Point For Health Monitoring Of Large Buildings

The pace of urbanization has become an unstoppable trend.The large buildings more and more appeared in people’s field of vision in the process.These buildings become an important symbol of modern cities.They are the main place where people study and work.These buildings usually have complex geometry structure and affected by internal and external factors.It inevitably contains the potential risk factors,which will bring great threat to the lives and property of the public security.Therefore,periodic comprehensive monitoring of large buildings is of great significanceWith the advantages of high precision,high efficiency and non-contact,3D lasers canning technology has been widely used in engineering safety monitoring.3D laser scanner can get a lot of high precision three-dimensional point cloud data in a short period of time.These huge Numbers of 3D point cloud contains a lot of redundant data,it will be great challenge for the point cloud data processing.Moreover,too much redundant points and mixed noise points have a negative impact on large buildings monitoring.This requests us that we should dilute the original data before we use them.Simplification process will destroy the geometric information in the cloud data,so how to reduce the amount of point cloud data at the same time keep the information as much as possible is a key.In order to solve these problems,this paper proposes a simplification method for large buildings three-dimensional point cloud data.We use Wuhan University of science and technology south lake new teaching building actual scanning point cloud as an example.We compiled our methods in MATLAB software platform and the PCL point cloud database.The main research content is as follows:(1)Geometric information identification of large buildings point cloud data.In order to keep the geometric information of the point cloud data,we should identification the geometry information from point cloud data before.On the basis of the tensor voting algorithm we proposed an adaptive tensor voting feature recognition method.Compared with the commonly methods,our methods can identification more abundant Geometry information and detail texture.Moreover,this method also avoids multi-scale calculation and complex calculation task.(2)simplification of 3D laser point data for large buildings health monitoring.The number of points cloud obtained by a 3D laser scanner is often very large,and the huge quantity will bring great challenge to the follow-up engineering deformation analysis.In order to make the point cloud data effectively and continuity,this paper puts forward a geometric features keep simplification method.Compared with the commercial software method our algorithm can more fully retain in the geometry information.Based on the actual large buildings 3D laser point cloud data experiment,we find that the proposed method is more accurate and efficient than the commonly way.It provides a scientific method for the safety situation of the special-shaped structures based on 3D laser scanning.Also it provides a highly efficient pre-processing method and has strong engineering practicability.