中文摘要：

针对视觉测量中椭圆目标的检测效率与定位精度较低的问题，提出一种高效的椭圆识别与定位方法。该方法利用回光反射控制点的强反射特性，提取目标的边缘信息，对连续的边缘进行分组，利用椭圆长轴信息对每组边缘数据进行椭圆识别，采用基于变量含误差模型的椭圆拟合方法，实现椭圆目标中心的精确定位。实验证明，该方法具有快速、自动化、定位精度高、鲁棒性好等优点，在视觉测量中具有广泛的应用前景。

英文摘要：

In order to solve the problems in 3D vision measurement such as low inspection efficiency and low location precision and so on,an efficient method of ellipse detection and location is proposed.Using high-reflective characteristic of the retro-reflective targets,the Canny detector is applied to the gray level image in order to obtain a binary edge-map.Edge contours points are extracted from the edge-map and divided into several groups according to the continuity.Ellipse objects can be recognized by using major axis information of an ellipse.An ellipse fitting method based on the Heteroscedastic Errors-In-Variable（HEIV） model is applied to realize the precision locating of the ellipse objects.Experimental results show that the method has such advantages as fast speed,automation,high precision and good robustness and can be extensively applied to the vision measurement.