Does particle filter really outperform low pass filter in indoor target tracking?
Daisuke Anzai, Shinsuke Hara · 2010
This paper compares the indoor tracking performance for a moving target between an IIR (Infinite Impulse Response) low pass filter and a particle filter. It reveals that, in the computer simulations where the motion of a target exactly follows the state transition model in the particle filter, the particle filter-based tracking system outperforms the IIR low pass filter-based tracking system. On the other hand, in the experiments where the motion of a target is really unpredictable and does not always follow the state transition model, the IIR low pass filter outperforms the particle filter. It concludes that the IIR low pass filter-based tracking system is advantageous in reality, because it requires far less computational cost and shows better tracking performance.