Investigation Of Rapid Quality Detecting Method From Large Steel Construction Based On T-LIDAR

Steel structures in recent years has been an unprecedented development, however, because of the weight of a single steel, strange shape and multi-part steel interface so its quality testing requirements and difficulty as great. Therefore, accurate measurement and testing is successfully installed the key steps of the steel-frame. In traditional GPS detection process, the presence of elements cover part of the design point does not exist, and therefore not all data points can be inquired, these data which for complex steel-structure are very limited. How fast quality inspection of large steel structures, from large steel structures in the international development trend point of view, there is much room for large-scale steel structures development in our country, so seek to rapid qualify detecting method to large-scale structure is very necessary.At the same time, three-dimensional laser radar technology in its data acquisition speed, high precision characteristics become a modern important measurement method. Today, laser radar technology is widely used in reverse engineering (or rapid prototyping), large industrial,3D modeling and other fields.3D laser point precision is to the millimeter level with the higher the accuracy of the model. Compared with traditional total station, GPS deformation monitoring methods, the collection efficiency of the three-dimensional laser radar sampling points, the formation of the entity in the data field, to avoid the limitations of the method of analysis point to an area. However, although the hardware and software to achieve the accurate measurement of the steel-frame structure. But rapid quality detecting has many deficiencies. In this context, this paper mainly from the following aspects:(1) Faced with a complex construction environment, combined with the actual project, writer talk about introducing three-dimensional laser for point cloud data of the whole process and to discuss the problems in the scanning process, how to quickly get high-precision point cloud data.(2) Point cloud of3D laser scanning technology to obtain a larger number of data in data processing often repeated traversal, in order to improve the operation efficiency were studied various indexes of the currently used for the operational characteristics to achieve athe linear octree create and the octree basis, interactive segmentation obtained data point cloud of the steel structure ends.(3) The corresponding normal vector of point cloud data, in combination with the type of steel-structure of the study, estimates the normal vector of a point cloud, and its success segmented plane point cloud.(4) Analysis of existing rapid quality detecting of large-scale steel-structure, summarizing its advantages and disadvantages, and based on the results of (3), the successful implementation of a steel-structure ends outside corner point coordinates for the outer corner point the coordinate precision of the corresponding analysis. Finally, by using the Visual Studio2010as the development platform, the paper talked about the design and the development the automatic extraction system of the steel-structure of outer corner point using OpenGL3D graphics library and C#programming language. It also talked about the realization of three-dimensional display, model scale, translation, single point selection as well as selection of multipoint-based interactive operation function of point cloud with scattered point cloud as the basic data format, realize a little cloud, and implemented the creation of linear octree, evaluation of normal vector, segmentation of plane point cloud, plane fitting,and square steel-structure frame splicing detection in the detection point (outside corner point) high-precision extraction as well as other basic functions through the system. Through experiments to analysis the accuracy of the feature points, at last according to application of the system of Tianjin West Railway Station construction project in the process of point cloud data processing, the results show that the accuracy and efficiency of the method has certain advantages.