For the NP-hard problem of track initiation for bearings-only multi-target tracking, an ant base rule based tracking
initiation method is proposed. Firstly, ant optimization algorithm is used to obtain the classification rule and further form the rule base. Then the rule base is employed to classify the candidate points, which are generated by the measurements from different sensors, and to extract candidates possible from real targets. Finally, the logic-based method is utilized to initiate tracks of interest among these extracted points. The numerical simulation results show the effectiveness of the proposed method, and the performance comparison of different methods is performed as well.