摘要
无人机场环境复杂。为了快速准确的找到无人机场的中心位置,首先对图像进行滤波等预处理,再根据二值图像特性以及椭圆几何特性,求取边缘跳变点的中点。利用所求跳变点的中点在竖直和水平方向的投影特点,以及改进的中心点估计法提取椭圆中心候选点。利用投影原理及跳变点特性对候选点进行筛选,分析得到机场中心点,最后利用边界追踪确认椭圆各个参数。实验数据表明,该方法具有很好的鲁棒性和精确性。
UAV airport is a complex environment. In order to find the center of airport quickly and accurately, firstly, pretreatment of image is carried on,such as filter. Secondly,according to the features of binary image and ellipse, the midpoint of the transition point of the edge was calculated. Using the projection characteristics of midpoint of the transition point ,which project in the vertical and horizontal direction, and the improved estimated methods, the center points of the candidate points are extracted. Using the projection principle and characteristics of transition point, airport center from the candidate points is selected.Finally,each parameter of elliptic through boundary-tracing is confirmed. Experimental data show that the method has good robustness and accuracy.
出处
《科技广场》
2009年第9期107-110,共4页
Science Mosaic
关键词
中心检测
几何特性
椭圆
Center Detection
Geometric Features
Ellipse.
