论文部分内容阅读
针对低信噪比条件下的机动小目标实时检测与跟踪困难的问题,提出一种基于混合估计多模粒子滤波的检测前跟踪改进算法.首先根据前一时刻所采用的模型状态及其转移概率等先验信息实现当前时刻的模型采样;然后在充分考虑当前量测下实现当前的粒子预测,采用一种序贯重要性平滑重采样策略,在不增加计算量的前提下,改善了粒子多样性衰退的问题;最后通过新的粒子集完成对模型和状态的合理估计与目标检测.仿真结果验证了该方法的检测与跟踪性能优于传统的多模粒子滤波方法.
In order to solve the problem of real-time detection and tracking of small maneuvering targets with low signal-to-noise ratio (SNR), a modified pre-tracking tracking algorithm based on hybrid multi-mode particle filter is proposed.Firstly, based on the model state and its transition probability And other prior information to achieve the current model sampling; and then take full account of the current measurement to achieve the current particle prediction, using a sequential importance of smooth resampling strategy, without increasing the amount of calculations under the premise of improving the diversity of particles Finally, a new set of particle swarm optimization is used to estimate the model and the state of the object and the target detection.The simulation results show that the proposed method has better detection and tracking performance than the traditional multi-mode particle filter.