针对复杂座舱环境设计了一个由inside-out(IO)光学跟踪器和outside-in(OI)光学跟踪器组成的新型座舱头部跟踪器.提出一种新颖的混合跟踪算法,利用扩展卡尔曼滤波器和融合滤波器融合座舱跟踪器中IO跟踪器和OI跟踪器的姿态信息,建立了实验装置验证提出的姿态融合算法.初步实验结果显示,与单一IO或OI跟踪器相比,该混合跟踪算法能够使座舱头部跟踪器获得更精确、更稳定、更鲁棒的位置方向信息.验证了所设计的座舱头部姿态跟踪器的结构合理性和所提出的混合跟踪算法的有效性.
The head pose tracker consisting of two complementary inside-out tracker and outside-in tracker is designed according to complex cockpit environment and then a novel hybrid tracking approach composing of two extended Kalman filters and one fusion filter is proposed to fuse the poses from the inside-out tracker and outside-in tracker.An experimental setup was established to simulate head pose tracker and verify the proposed hybrid tracking approach.The primary experimental results show that our head pose tracker is capable of achieving more accurate and stable data of position and orientation as well as better robustness compared with using single inside-out or outside-in tracker.It also validates the rationality of the tracker configuration and the usefulness of our hybrid tracking approach.