In order to match the intersectant targets' tab when these objects separate, a kind of method that makes use of fuzzy multi-parameter to match targets is proposed. Multi-parameters include the forecasted position after targets separate, target speed, target size and geometry shape, and different parameter corresponds to different influence coefficient, evaluate the integrated matching coefficient that every observation target corresponds to forecasted target, and the corresponding relationship that matching coefficient is the biggest is the best matching result. In order to locate the position and speed of forecasted target, a new method of track forecast is proposed. firstly, Hough transform is used to the object's track before objects intersect, by this, the object's positions that the warp between forecasted location and observation location is large enough is eliminated, then least square method is used for track forecast by the remainder valid positions, and get hold of the forecasted position when objects separate. The experiment results show: when tracking the bulky objects which intersect and separate, the right identification probability of traditional least square method is 87.5%, and the right identification probability of fuzzy match by multi-parameters can attain 96%, the reliability improves in evidence.