The non-negative matrix factorization (NMF) is widly used in data fusion for the advantage of feature extraction

and the particle filter (PF) is an effective method for the state estimation of non-linear and non-Gaussian dynamic systems.Therefore

an infrared small target tracking algorithm based on improved particle filter is proposed.Current and previous particle distribute weights are fused by NMF in order to reduce the precision error caused by particle divergence in classic PF method.So the tracking error of sheltered and disappeared target can be avoided.Experimental results show that the proposed method has better tracking precision and is more stability for small target tracking than the classic PF method.