一种基于粒子滤波的多声源跟踪算法
发布时间：2019-06-07 20:29
【摘要】：研究了声源轨迹交叉情况下的两个声源跟踪问题。针对传统跟踪算法容易丢失较弱声源的缺点,提出一种基于粒子滤波的改进多声源跟踪算法。该算法在跟踪过程中估计声源之间的距离,当此距离小于0.3 m时,根据声源粒子的速度方向来调整粒子权重,引导粒子向各自声源的运动方向移动,从而避免较弱声源的粒子被较强声源的粒子"吸引",保持了粒子的多样性。仿真结果表明,在不同的交叉角度下,相比于传统算法,改进算法能大幅度减小对较弱声源的跟踪丢失率,并能有效提高两个声源的跟踪精度。
[Abstract]:Two sound source tracking problems under the condition of cross sound source trajectory are studied. In order to solve the problem that the traditional tracking algorithm is easy to lose weak sound sources, an improved multi-source tracking algorithm based on particle filter is proposed. The algorithm estimates the distance between sound sources in the tracking process. When the distance is less than 0.3 m, the particle weight is adjusted according to the velocity direction of the sound source particles, and the particles are guided to move in the motion direction of their respective sound sources. Thus, the particles with weak sound source are "attracted" by the particles with stronger sound source, and the diversity of particles is maintained. The simulation results show that compared with the traditional algorithm, the improved algorithm can greatly reduce the tracking loss rate of the weaker sound source and effectively improve the tracking accuracy of the two sound sources at different cross angles.
【作者单位】： 九江职业技术学院电气工程学院;东南大学信息科学与工程学院;
【基金】：国家自然科学基金青年科学基金资助项目(61201345)
【分类号】：TN912.3
[Abstract]:Two sound source tracking problems under the condition of cross sound source trajectory are studied. In order to solve the problem that the traditional tracking algorithm is easy to lose weak sound sources, an improved multi-source tracking algorithm based on particle filter is proposed. The algorithm estimates the distance between sound sources in the tracking process. When the distance is less than 0.3 m, the particle weight is adjusted according to the velocity direction of the sound source particles, and the particles are guided to move in the motion direction of their respective sound sources. Thus, the particles with weak sound source are "attracted" by the particles with stronger sound source, and the diversity of particles is maintained. The simulation results show that compared with the traditional algorithm, the improved algorithm can greatly reduce the tracking loss rate of the weaker sound source and effectively improve the tracking accuracy of the two sound sources at different cross angles.
【作者单位】： 九江职业技术学院电气工程学院;东南大学信息科学与工程学院;
【基金】：国家自然科学基金青年科学基金资助项目(61201345)
【分类号】：TN912.3
【参考文献】
相关期刊论文 前3条
1 金乃高;殷福亮;陈U，
本文编号：2495063
本文链接：https://www.wllwen.com/kejilunwen/wltx/2495063.html
