SHAPE RECONGNITION IN MACHINING PROCESS

加工过程中的形状识别是实现自动化制造系统的一个重要问题。在机械加工系统中，所需的加工形状是由几何模型给出的，基于几何模型的计算机视觉系统适合于零件形状识别系统，本文的主要研究工作如下：1）建立图像点与几何模型点之间的映射过程。- 介绍DKA方法用于映射过程。2）提出了一种新的求解过程模型--视觉协议的概念--视觉协议算法的开发3）通过计算机仿真验证了所提出的方法，得到了以下结果：1）DKA方法是解决计算机视觉中P4P问题的有效方法。2)小位移旋量法是解决计算机视觉中多点映射问题的有效方法。3)认为所提出的视觉协议模型包含了手工编译方法、解释器方法和编译方法，是一个通用的形状识别过程模型。4)用C^<++>计算机编程语言实现了所提出的视觉协议。

It is generally recongnized that shape recongition in machining process is a very important issue for realizing an automated manufacturing system. In the machining system, required machined shape is given by geometrical model, then a model based computer vision system will be adequate for the part shape recognition system.Aims of research works are follows : 1) To establish the mapping procedure between the points on image and points of geomarical model. -introduce the DKA method for mapping procedure. -introduce the Small Displacement Screw method for mapping proceduure 2) To propose the reconginitive process model -introduce the Vision Protocol concept -development the algorithm for vision Protocol 3) Verification of the proposed method by computer simulationFollowing results are obtained : 1) It is recognized that DKA method is very effective for P4P problem in computer vision. 2) It is recognized that Small Displacement Screw method is very effective for multi point mapping problem in computer vision. 3) It is recongnized that proposed Vision Protocol Model contains the hand-complie method, interpretor method and compile method, and is a general shape recognition process model. 4) It is recognized to realize the proposed Vision Protocol by using C^<++> computer programming language.