This paper analyses the PDA algorithm, which is commonly used in plot-track association. Aimed at the problem of bug tracking when it works in dense targets environment, an improved algorithm is proposed and the concept of subjection degree is introduced. After determining the subjection degrees between all the effective measurements and the corresponding targets by calculating statistical distance, intersecting trace windows are split, then PDA algorithm is applied to associating plots with tracks in independent track windows, and Kalman filter is applied to estimating real states of targets. In the simulation, the algorithm proposed in this paper is compared with JPDA algorithm, and the results show the proposed algorithm is slightly decreased in correct association rate, and substantially decreased in calculation.