Abstract: A locational lot number association algorithm is proposed for joint multi-station observation of non-cooperative target localization scenarios. The algorithm uses fuzzy theory to describe the fuzzy membership function between the direction information, continuous spectral feature information, and line spectral feature information in the correlated set and the measure to be correlated. A comprehensive membership matrix of fuzzy is constructed for the two-dimensional association, reducing the impact of low membership by setting a membership threshold. Simulation results show that the algorithm overcomes the problem that the success rate of association decreases when the number of targets increases, and the recall rate of association results can reach more than 0.8 when there are certain missed and false alarms, and the accuracy of association results is also maintained at a high level. Both simulation results and experimental data confirm the superior performance of this method for target correlation localization in complex underwater environments.