基于矢量约束的随机特征点选取算法

无人机利用视觉在未知区域自主着降时,提取的特征点具有数量多、随机性强等特点。针对利用随机特征点进行位姿估计精度低、稳定性差的问题,提出一种基于矢量约束的随机特征点选取算法。首先通过分析位姿估计方程可知，特征点地理坐标是影响方程组求解精度的重要因素;然后在引入矢量角均分度、矢量模值均值及矢量模值最大值三项约束指标基础上，制定了一种基于矢量约束的特征点选取策略;最后利用正交迭代算法对所选取的特征点进行位姿估计精度验证。实验结果表明，提出的算法相比于随机选取的特征点进行相对位姿估计精度更高,鲁棒性更强。

Selection algorithm of random feature points based on vector constraints

The feature points of the extracted image are characterized by a large number and strong randomness when unmanned aerial vehicle (UAV) landing autonomously at an unknown zone by using vision. In order to overcome the problems that randomly selecting feature points for relative position and angle estimation leads to low precision estimation and poor stability, a selection algorithm of random feature points base on vector constraints is proposed. Firstly, geographic coordinates of the feature points are considered as an important factor which affects the equation precision through analyzing the position and attitude estimation equation. Secondly, the vector angle average degree, the mean of vector modulus and the maximum value of vector modulus, three kinds of constraint functions are introduced. And a selection strategy of random feature points based on vector constraints is developed. Finaly, the orthogonal iterative algorithm is used to evaluate the position and attitude estimation accuracy for the selected feature points. The experimental results show that the proposed algorithm has higher accuracy and stronger robustness compared to the method of randomly selecting feature points.