Marine Target Magnetic Anomaly Detection Based on Multitask Deep Transfer Learning
Shigang Wang,Wenhua Song,Xiangyuan Zhang,Yaqiu Qin,Bin Li
DOI: https://doi.org/10.1109/LGRS.2023.3273722
IF: 5.343
IEEE Geoscience and Remote Sensing Letters
Abstract:As an important physical field feature, the magnetic anomaly is widely used in the detection of marine ferromagnetic targets. Influenced by the complex ocean measurement environment, the collected magnetic data are usually contaminated with noise, and traditional magnetic anomaly detection methods are less versatile due to the specific conditions required. Faced with the challenging situations in the ocean, the use of deep learning methods can not only automatically extract discriminative features from the magnetic data but also reduce manual interference in the detection model design. Deep learning requires a large amount of data to train the model, but the real target magnetic anomaly data is expensive to acquire and small in number. In this letter, we propose a multitask deep transfer learning model for simultaneous denoising and detection of marine target magnetic anomaly data under limited labeled samples. Specifically, a convolutional denoising auto-encoder (CDAE) network is designed for adaptive background noise modeling and discriminative feature learning. Meanwhile, a fully connected classification (FCC) network is cascaded and jointly trained with it for simultaneous noise filtering and anomaly detection. To guarantee sufficient learning effect, both real measured data from sea trials and simulated data from the magnetic dipole model are used for training the network in a two-stage manner. During the experiments, the learned multitask deep model is, respectively, used to denoise the marine magnetic data and detect target anomaly signals. It achieves both superior noise-removal and anomaly discovery performance than the traditional methods, and meanwhile, is much more efficient.
Environmental Science,Physics,Engineering,Computer Science