机器视觉二维反求测量系统的研究

对机器视觉二维反求测量系统进行了研究，以实现二维复杂平面边缘尺寸的数字化录入；采用虚拟网格技术实现大场景下的测量标定；利用双CCD分区域获取图象，通过设置规则参照物、数据拼接等数据处理提高测量精度；提出了利用8～链码的“链差”确定转折点，用“变长度虚拟直线”判断线型实现图形矢量化的方法;数据输出采用通用的DXF文件格式实现与CAD软件的接口;研究结果表明，该系统能够应用于二维复杂平面边缘的反求测量。

2-D Measurement System for Collecting Object Edge Data Based on Machine Vision

A reverse measurement system about the data collecting of 2-D object edge based on machine vision is discussed, in order to realize digital data input of the dimensions of 2-D object complex edge. This system realized calibration in large scene by virtual grid calibration method, and advanced the precision of system by using double CCD cameras, it realized data mergence by setting up referencer of regular shape. Some methods were put forward to realize graphics vectorization such as using eight-chain code difference to search for the turning points, and using change-length subjunctive beeline to distinguish linetype. The output data use the general DXF document format to connect the software of CAD. Research indicated that this system can be applied in reverse measurement to the dimensions of 2-D object complex edge.