基于光学相机的光流-相机定标融合定位算法

Optical Flow-Camera Calibration Fusion Positioning Algorithm Based on Optical Camera

为了解决光流算法在相机大转角情况下定位误差过大和相机定标定位算法在LED数量不足的情况下无法定位的问题,提出将相机定标定位算法和光流算法进行卡尔曼滤波融合的光流-相机定标融合定位算法。首先,采用相机定标定位算法进行室内定位;其次,引入光流算法补偿相机定标定位算法的计算结果;最后,采用卡尔曼滤波融合2种算法的定位结果。搭建实验平台验证所提算法的定位性能。实验结果表明,所提算法能够解决相机定标定位算法在LED数量不足的情况下无法定位的问题,同时能够克服光流算法在相机大转角情况下定位误差过大的问题,将相机在大转角运动情况下平均定位误差从6.86 cm降低至1.02 cm。

In order to solve the problem that the positioning error of the optical flow algorithm is too large in the case of large camera rotation angle and camera calibration positioning algorithm is unable to locate when the number of LEDs is insufficient,an optical flow-camera calibration fusion positioning algorithm that combines the camera calibration positioning algorithm and optical flow algorithm with Kalman filtering is proposed.In particular,the camera calibration and positioning algorithm is used for indoor positioning.Then the optical flow algorithm is introduced to compensate the calculation results of the camera calibration and positioning algorithm.Finally,the Kalman filter is used to fuse the positioning results of the two algorithms.An experimental platform is built to verify the positioning performance of the optical flow-camera calibration fusion positioning algorithm.The experimental results show that the proposed algorithm can solve the problem that the camera calibration positioning algorithm can not locate when the number of LEDs is insufficient,and can reduce positioning error of the optical flow algorithm in the case of large rotation angles.Specifically,the average positioning error of the camera at large rotation angles has been reduced from 6.86 cm to 1.01 cm.