A distributed target tracking algorithm is proposed for the data packet dropouts and missing measurement in distributed target tracking process. The algorithm is composed of a local filter under statistics and a fusion filter based on covariance crossover algorithm. At first, the local filter used neighborhood node to measure information and calculate local filter value, adopted fusion filter to carry out fusion processing of neighborhood node local filter value, and acquired node target coordinate information. At last, use typical target tracking model to carry out simulation analysis for the algorithm. The simulation model showed that, the method can effectively restrain the bad influence of incomplete information on target tracking result. It can provide reference for design and research of distributed target tracking system in practical engineering.