The Study Of Surface Reconstruction Technologies For Reverse Engineering And Web-based Application System Development

AbstractSudece reconstruction, aiming to achieve the mathematical expression of geometricmodel for complex free-form surface from the scattered measured dat4 is a key techniqueof Reverse Engineering. Suffoce reconstruction software plays an important role to reducethe task of designer, reduce designing time, improve designing efficiency and improveproduct quality.With the poPularization of lntenert and the rapid development of Web technique,profound innovations are tuming up in the area of the traditional product design,manufactUring and producton mode. Consequently a series of new concepts and newtecImiques concerning collaborative design and manufacturing, geographical distributeddesign and manufactUring and global manufacturing are coming out. The application ofWeb technique into Reverse Engineering System will bring vigor to Reverse EngineeringSystem. Hence, it is prospeCtive and promising... The dissertation presents a research and disscusion fOr the technique of complex free-form surface reconstruction in Reverse Engineering. Serveral kinds of algorithm fOrsurface reconstruction were presented with different basis functions. The dissertation focuson the algorithm for the reconstruCtion of B-Spline Suffece from the scallered measureddata defined on the reCtangle domain or with arbitrary topological shape. Based on the -above theory reasearcly we combined suthee reconStruedon system developmentteclmology with Web tedrique, fuwher stUdied the key technique of Web-based suffocereconstructon system (WebSurafcer) development, and conancted a prototype system ofWebSurfacer, and finally discussed the application of WebSuffocerThe main points of the dissertation are the followings;1. A tWo-steP linear aPproch is emPloyed for surface fittillg from scattered data withrectangle topological shaPe. It is very important in the aPproch to evalute the boundarycondition. TWo aPprochs to evaluating boundary condition are presented in thisdissertation. They are call m-relationship solution method and Bessel method.2. The method of least square B-Spline surface fitting to scanered data is presented. Inthis method, a base sudece is constructed firstly, the paramCterization is realized byprojecting the measured points to the base surface. The parameters of the projectedpoints are then used as the parameters of the measured points. For cosmically scattereddata, we use normal equation method to solve equation syStem. Concurrently, energyfunction is used as a constralnt. The aPproximation is a multi-step iteration process.Finally, a smooth suffoce is achieved. The aPplication result shows that this aPprochcan handle super suface, and the computation cost is reduced.3. A new surface fitting method is disscused. The method is a adaPtive process to fit thedata by using radial basis functions in least square mean. An aPproch tO find theoptimal parameters value is presented. It is showed that radial basis fonctions surfacemis superior to B-Spline surface in that it is easy to achieve smooth surface by using radial basis function, and generally, to express a surface using radial basis function needs fewer control points than B-Spline.4.To the scattered data set of topological type, a procedure for reconstruction of BSpline surface is presented too. Firstly, in the procedure, a rectangular mesh M~ is found and all the scattered measured data are projected ontoM0. Secondly, using every rectangle as domain, projection values as parameters value, a B-Spline Surface consiting of bi-cubic Bezier surfaces merged with hi-quadratic Bezier surfaces is reconstructed on the domain on the constraint that all the patch are merged oncontinuity.5.With the intergration of Web technique and surface reconstruction system development technology, a Web-based surface reconstruction system WeSurfacer was developed. Some good application results were achieved.