Research On Measurement Planning And Model Constructing Of Free Surface In Reverse Engineering

Reverse Engineering (RE) is the reverse procedure of product design and manufacturing, which is different from the traditional one. It is also a practical technology in developing new products. Data acquisition and point cloud pre-processing are two key links of RE. Based on 3D digital laser measurement system, data measurement, data preprocessing and surface reconstruction are systematically studied. The results have an important meaning on product modeling and subsequent machining.Based on the principle of 3D laser measurement, data acquisition was studied on integrity and accuracy of data measuring through a great amount of experiments. According to the characteristic of measuring equipment and complexity of object contour, the measurement planning in different conditions was proposed and the correctness of measurement planning was verified by experiments; through orthogonal test, the influence of LSH800 parameters setting for measurement accuracy was analyzed; analyzing surface characteristic of test workpiece (such as surface color, inclination, surface curvature, luster,etc.), the effect of different surface character on measuring data was obtained. Experiment results show that in order to get perfect point cloud, the reasonable measurement planning should be established, the appropriate measuring parameter should be choosed and necessary surface preparation should be done.Several problems in the data processing of non-contacted 3D scanning measurement were discussed in this thesis, such as noise points deleting, data simplification, mooth processing as well as point cloud data registration and integration on the basis of analysis on existing means of reducing point cloud, the method based on octree space division is proposed to reduce points cloud, while the geometric characteristic of raw data is kept in a better mode. Scanned data from different view is transformed by superposition of three datum points, graphic data is fine transformed through ICP algorithm and Its Application; for the characteristic of linear scanning data, feature extraction method based on linear scanning data is given.In this thesis, NUBBS is choosed as a standard form of free curves and surfaces through research on properties and mathematical model of the curved surface. The regularity and methods of surface reconstruction from point cloud data are summarized and reconstruction methods centered at feature line are acheieved. Finally, a numerical example shows that the method is feasible and effective.