增强现实中的混合跟踪算法研究

单一的跟踪方法存在较大的局限性，为提高增强现实中跟踪环节的实时性和准确性，针对视觉跟踪和磁力跟踪的特点进行研究，提出了一种基于自适应粒子滤波的混合跟踪算法，用于对头部运动轨迹估计。该算法通过分析系统状态，自适应地融合多传感器数据，并建立相应的状态转移模型和系统量测模型；另外，该算法能在非线性非高斯的环境下动态地改变滤波器的粒子数和噪声方差，最终实现对头部运动轨迹的实时、正确估计。实验结果表明，该算法能有效地提高基于视觉和磁的混合跟踪的鲁棒性和运动估计的准确性。

Research on the Hybrid Tracking Algorithm Based on Self-adaptive Particle Filter

Single tracking method limites the real-time performance and accuracy of the tracking process in Augmented Reality. This paper proposes a hybrid tracking algorithm based on self-adaptive particle filter, which is designed for the estimation of the head motion. By analyzing the system status,the algorithm fuses data from multiple sensors self-adaptively and builds the corresponding state transition model as well as system measurement model.The algorithm fulfills real-time and accurate estimation by varying the quantity of particles in the filter and the noise variance. The experimental results show the algorithm can effectively improve the robustness for the hybrid tracking and accuracy of the motion estimation.