一种基于Hough变换新的亚像素定位算法  

The new sub-pixel location algorithm based on Hough transform

作　　者：陈光静[1] 陈明贵[1] 肖玉强[2] 

机构地区：[1]西安通信学院电子工程系,陕西西安710106 [2]西安电子科技大学机电工程学院,陕西西安710071

出　　处：《电子设计工程》2010年第12期127-130,共4页Electronic Design Engineering

摘　　要：为了提高数字摄影测量的精度,提出了一种基于Hough变换的亚像素定位算法,针对直线边缘检测和圆边缘检测,分别采用二次Hough变换法和三等分Hough变换法,通过使用二次Hough变换的方法,克服了直线边缘计算精度低和速度慢的问题,通过使用三等分Hough变换法,克服了传统Hough变换法圆边缘检测性能受参数离散间隔制约及精度低的问题。仿真和实验表明新算法的亚像素定位精度能达到0.02像素,有效地提高了数字摄影测量的精度。In order to improve the accuracy of digital photogrammetry, a sub-pixel location algorithm based on Hough transform was put forward for straight edge detection and circle edge detection, adopting the second Hough transform method and the trisection Hough transform method respectively. By using the second Hough transform method, the problem that computing straight edge was low in precision and slow in speed was overcame. By using the trisection Hough transform, the problem that the detection performance of the circle edge by using the traditional Hough transform could be constrained by the parameters of discrete interval and was low in precision. Simulation and experiment show that the new algorithm can make the sub-pixel positioning accuracy within 0.02 pixel and effectively improve the measurement accuracy of digital photog-raphy.

关 键 词：数字摄影测量 HOUGH变换 亚像素定位 直线边缘检测 圆边缘检测 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]