ENGINEERING RESEARCH EQUIPMENT GRANT: Robot Vision System for Advanced Industrial Automation

一个机器视觉系统正在获得，它将用于支持目前在三维物体识别和图像分析方面的研究。该系统将被用来探索一种新的、通用的方法来识别、描述和理解任意方向的三维物体，给出一个单一的二维相机图像和预定义的物体模型。该模型由三维空间中的一系列边向量和顶点组成。用于检测对象及其取向的过程包括将来自模型的边缘对匹配到图像中的边缘对。利用匹配的数据，可以开发一个专家系统，将信息汇编成对3-D场景的全面描述集。

A machine vision system is being acquired, which will be used to support current research in three-dimensional object recognition and imag analysis. The system will be used to explore a new, general approach to recognizing, describing, and understanding three-dimensional objects in any arbitrary orientation, given a single 2_d camera image and predefined object models. The models consist of a series of edge vectors and vertices in 3-space. The procedure for detecting an object and its orientation involves matching of pairs of edges from the models to pairs of edges in the image. Using the matched data, an expert system can be developed to compile the information into a comprehensive set of descriptions of the 3-D scene.