Non-contact Scanning Measurement For Three-dimensional Profile And Research On Reverse Engineering

The precision measurement is a problem which has a wide range of application and greatly practical value. With the development needs of measurement calibration and the limits of contact measurement, other new non-contact measurement technologies such as image measurement and laser measurement have been developed greatly. This paper is financed by the National Defense project for the 11th Five-Year Plan (B20301118). The aim of the project is to build a new type of composite measurement system. This paper mainly studies on 3-D scanning measurement of laser measure system and the reverse engineering techniques base on scanning data.This paper mainly researches the laser triangulation and the application of three-coordinate measurement machine (CMM) in three-dimensional profile (3-D profile) measurement, a 3-D laser scanning measuring system which integrated the laser triangulation and CMM is designed. The main purpose of this research is to realize scanning measurement of the 3-D profile, a laser triangulation measurement system which is integrated in the CMM is designed, and software measurement system establishment is also completed. Besides, this paper give systematic studies on data acquisition, the point cloud data pretreatment technology, and curve, surface reconstruction technology and so on. All this has a significant influence on the product modeling and new product development.This article introduces the laser measurement system devices, and completes the software platform structures of the measurement system on account of the target and the function demands of this system. According to the measurement requirements, this paper plans the main flow and each functional module of the system. Not only realizes scanning measurement, equipment management, motion control, database management and other basic functions in software, but also develops graphical display function of the system using OpenGL, and realizes 3-D graphic display and operation of point cloud data.Debugging tests have been conducted on the measurement system. The influence on measurement results caused by the object surface properties such as surface colour, surface slope, surface gloss has been taken out. Through in-depth analysis, the principle of the measurement error caused by the surface gradient is obtained. Then use the main conditions of the measurement errors as input parameters to build the neural network model according to which realizes error compensation.Finally, this paper researches some point cloud pretreatment methods such as the point cloud denoising, streamlining point cloud, point cloud patches and so on, and curve, surface reconstruction in reverse engineering. Some point cloud pretreatment methods has been introduced simply. And several common curves, surface reconstruction methods such as Bezier, B-Spline, NURBS, surface reconstruction are analyzed, then compares three different curves, surface models. Using the software of ImageWare realizes the point cloud pretreatment and curve, surface reconstruction.