Research On Surface Interrogation Methods Based On Illumination Model

With the development of laser scanning and rapid prototyping technologies, reverse engineering is widely used in automotive, aviation, mold and other industrial fields. At the same time the smoothness of product modeling surfaces is increasingly important, which makes surface interrogation become an essential part of high-quality reverse engineering. Aiming at several key problems of surface interrogation methods based on illumination model, the research is concentrated on improving calculation model, increasing analysis efficiency and realizing quantitative description of surface fairness.Continuous surface interrogation technologies and several optimization algorithms based on illumination model are studied. By optimizing distance function model of reflection lines, the problems of distortion of reflection lines and strong correlation of reference light source are solved. The regeneration efficiency of dynamic highlight lines has been improved by updating local distance surface when surface is locally deformed. As the lack of information provided by highlight lines, highlight curvature line method is put forward. The calculation models of circular highlight lines and isophotes are optimized for inmproving the computational efficiency.Quantitative analysis of surface fairness and optimized algorithms of highlight line with high-precision are studied. Energy density model and energy square method are proposed based on curve energy, which demonstrates the feasibility and practicality of energy method on quantitative analysis of surface fairness. Optimized algorithm of highlight points is provided by combining directional derivative and Newton iterative algorithm, and the high stability, fast convergence and high efficiency of the optimization algorithm is verified, which ensures the smooth progress of the quantitative analysis.Discrete surface interrogation technologies based on illumination model are studied. Feature extraction algorithm based on tensors voting method is studied to detect one-order continuous feature. Then the extraction method of generalized mesh boundary is put forward. Thus surface interrogation methods can be used to check both the continuity and fairness of the discrete surfaces.Software platforms of surface interrogation on both continuous and discrete surfaces are completed. With the quadratic exploitation tools based on UG, the software of continuous suface interrogation is developed, and then quantitative analysis and error detection of curves are realized. Surface interrogation on discrete surfaces and algorithms validation are realized by combination Visual C++ and OpenGL graphic technology.On the constructed software platforms, quality analyses of styling surfaces such as automobile covering parts and turbine blades are completed, the effectiveness and feasibility of the optimized algorithms and analysis methods are verified successfully.