基于视觉的自由曲面快速测量方法技术

In this project, the high resolution digital camera is used to obtain the surface image of the parts from different directions，and a measuring method of free-form surface, the method of shape from shading, is presented. Through analyzing the geometric error of the surface reconstructed by using four conventional local illumination models, such as Lambert, Phong, Torrance-Sparrow and Cook-Torrance model, an improved illumination model suited for the object measurement is derived. Based on the improved illumination model, a numerical solution of the image irradiance equation is proposed by using theory of nonlinear partial differential equation and improved according to the experiment and calculation results of the basic objects. Further more, feasibility of applying the method of shape from shading in non-contact measuring free-form surface is analyzed. The algorithms of reconstructing surface based on the monocular single image and monocular multiple images are studied respectively. Simultaneously, many basic objects made of the same material as the parts are measured by the method of shape from shading in order to calibrate the parameters of the illumination model. At last, the measuring method is applied to the RP manufacturing and CNC profile milling and provided to be effective and feasible.

以光度学、图象处理及非线性偏微分方程的理论与方法的基础、高分辩率数字相机为测量手段、研究通过综合分析从不同角度拍摄的自由曲面灰度图象的明暗关系，而获取自由曲面三维几何形状的理论与数值方法，发展出自由曲面的快速非接触测量方法与技术，具有对测量环境要求低、工作距离远，操作方便、并能快速获取被测曲面大量信息等诸多优点。.