In this paper，the dipole magnetic anomaly detection model is developed to determine the underwater unexploded ordnance（UXO）. The weak magnetic signal is extracted from the acquired magnetic signal using the orthogonal basis functions （OBFs）. In combination with GPS longitude and latitude information，the data fusion process is proposed for processing seabed magnetic anomaly data，and then a geomagnetic map is created to locate the suspicious UXOs. The algorithmic processing flow is analyzed by simulation. Firstly，a finite element simulation method is used to model the magnetic anomaly generated by the buried objects under the geomagnetic background. And then the actual spatial sampling process is simulated to obtain the sampling signal map of the observation area. The magnetic field is reconstructed by interpolation method，and finally the location information of the anomaly source is obtained. In practical application，the SNR of the original magnetic anomaly signal is 14.34 dB. After filtering and OBFs signal processing，the SNR is enhanced to be 20.04 dB，which is significantly increased by 5.7 dB. Finally，the geomagnetic map is used to determine the longitude and latitude position of the unexploded ordance that may exist. The result shows that the accuracy of target detection reaches 100% and the false alarm rate is zero. This study verifies the accuracy and reliability of this proposed magnetic anomaly detection and localization method for buried UXOs in shallow water.
