摘要
微小型飞行器飞行高度的实时测量是实现其自主飞行的关键技术之一。为了准确快速地获取微小型飞行器的飞行高度参数,提出了一种新的稳健的利用图像信息获取微小型飞行器飞行高度的方法。此算法首先使用光学理论和几何学理论推导出了飞行高度的计算公式,综合运用Harris角点检测算法和区域灰度算法得到图像间的匹配角点,并使用分割区域法得到图像之间的关键点。最后通过计算关键点在图像中的位移情况,计算出飞行器当前的高度值。飞行实验表明,这种测量算法可以实时获得飞行器的高度,精度满足飞行器平稳飞行的需要。
Flight height is one of key technologies for micro aerial vehicle （MAV） flight autonomy.In order to get the MAV flight height accurately and quickly,a new robust flight height measurement algorithm was proposed.In the algorithm,the height was deduced from optical theory and geometry theory,and a matching corner points set was obtained by Harris corner detector and regional gray method.And then,the key points were got using the segmentation area method.In the end,the current flight height was obtained by computing the displacement of the key points in the images.The flight experiment shows that flight height measurement is real-time,and it meets the stable flight control requirements for the MAVs.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2010年第12期2784-2787,共4页
Journal of System Simulation
基金
国防科工委基础科研基金(A2220060045)
关键词
微小型飞行器
角点检测
角点匹配
参数提取
MAV
corner points detector
corner points match
parameter estimation
