二次曲线图像自动鲁棒提取研究  

Study on Automatic and Robust Extraction of Imaged Conics

作　　者：吴劭桓 郭世毅 WU Shaohuan(Taiyuan No.12 Middle School, Taiyuan 030002, China;NLPR, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China)

机构地区：[1]太原市第十二中学,山西太原030002 [2]中国科学院自动化研究所模式识别国家重点实验室,北京100190

出　　处：《电子工业专用设备》2019年第1期66-71,共6页Equipment for Electronic Products Manufacturing

摘　　要：对图像中的二次曲线进行自动提取有着广泛的应用价值,可用于相机参数的标定、机器人定位、增强现实等众多领域。目前已有的方法通常是对单个的二次曲线进行提取,并且在单个二次曲线提取中,当二次曲线不完备时,提取通常会失效。对此提出了一种对多个二次曲线图像自动拟合的算法,创新性包括:(1)提出一种将高斯消元与Hough变换相结合的二次曲线检测方法;(2)提出将RANSAC与Hough变换相结合进行二次曲线检测;(3)应用一种新的几何距离进行二次曲线拟合。该方法可以处理数据不完备或者冗余的情况,不仅适用于单个二次曲线自动拟合,也适用于多个二次曲线拟合。与已有方法相比,实验结果表明该方法的高效性。Automatic extraction of imaged conics has wide applications such as for camera calibration, robot localization, and augmented reality. Most of existing methods for extracting conics are only for a single conic and are not robust to occlusions. In this paper, we study automatic extraction of multi conics in images, where the process includes not only detection but also fitting. The novel contributions are as follows: 1) We propose a fusion strategy of Gaussian elimination and Hough transformation for conic detection;2) We propose a combination of RANSAC and Hough transformation for conic detection;3) We apply a new geometric distance to conic fitting. The proposed method can deal with not only redundant but also incomplete data. It is suitable to both single and multiple conics. Experiments including comparisons show the effectiveness of the method.

关 键 词：二次曲线拟合 二次曲线检测 霍夫变换 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]