It is hard to detect and track special targets with super high-speed and super strong- maneuverability efficiently, only by traditional sensors such as radar. Therefore, intelligence information and expertise base etc. should be needed. Nevertheless, there are no practicable means for the management of multi-source heterogeneous information registration. The characteristics of special targets are introduced and analyzed firstly. Some kinds of heterogeneous information are expressed in Finite-Set Statistics(FISST) theory and some corresponding filtering algorithms are put forward. The matching probability of focused targets is taken as the registration index of heterogeneous information, and the results are computed with Kalman evidence filter. Numerical simulation validates the feasibility of the proposed method.