视觉辅助下的室内惯导位姿修正

Pose correction of INS for indoor location based on a vision navigation method

惯性导航系统可以短期内提供连续的高精度信息,但是误差会随时间增大,不能长期独立工作。而在大型仓库、地下停车场等室内卫星信号薄弱的场景中,传统的惯导+卫星组合方法也不再适用。针对该问题,本文提出了一种视觉与惯导组合定位的方法。本文研究的惯导+视觉组合的定位方法中,采用基于合作目标的单目视觉定位方法对惯导误差进行修正。对于惯导误差的修正方法,本文利用视觉定位的位姿信息建立量测方程,进行卡尔曼滤波,并选取合适的试验设备,通过实际试验对比验证了该算法对惯导系统误差的修正具有良好的效果。

The inertial navigation system can provide continuous navigation information,of which short-term positioning accuracy and stability is good. But positioning error increases with time which cannot work independently for a long time. The traditional way is to use the INS+GNSS integrated navigation method to correct the INS cumulative error. However,this method is no longer applicable in the situation where satellite signals such as indoors and underground are weak. In this paper,vision navigation method adopts a monocular vision positioning method based on cooperative target to correct the error of inertial navigation system. For the method of correcting the error of inertial navigation system,this paper uses the position and attitude information of vision navigation system to establish the observation equation,realize the Kalman filter algorithm based on vision correction,and show that the algorithm has good effect on the correction of the error of inertial navigation system through experiment.