Magnetic anomaly detection techniques have been extensively utilized in civil and military applications. One of the most prevalent detection methods is known as the orthogonal basis functions, which is not satisfactory in colored background noise and low signal-to-noise ratios. In this paper, a residual network based on convolutional block attention module (CBAM) is proposed. The input to the network is the time-frequency spectrum obtained by wavelet transform of the raw magnetic data. The residual modules help the network to learn features better and improve performance, while the CBAM facilitates focusing on important features of the data. Experimental results show that the detection performance of the proposed network has better detection performance than some existing methods at low signal-to-noise ratios.

This paper is concerned with the aeromagnetic compensation, which extracts and subtracts the disturbing magnetic fields generated by the aircraft itself. It is a crucial step in aeromagnetic anomaly target detection. The classical Tolles-Lawson model is established based on many assumptions, including at least the following three points. First, the nonorthogonality, zero offsets and nonlinearity of the triaxial fluxgate are ignorable. Second, the aircraft has no movable part and keeps rigidity during flying. Third, the ambient electric field does not induce current on the aircraft surface to produce extra magnetic interference. By taking account of these factors, enhanced models are derived in this work. Meanwhile, fitting and filtering schemes are introduced to suppress the impacts of geomagnetic gradient and diurnal magnetic variations. Field experimental data are used to examine the enhanced models and algorithms. Compared with the original Tolles-Lawson model, the proposed models can significantly improve the compensation accuracy and signal-to-noise ratio, which facilitates the subsequent detection for a magnetic anomaly target.

In this letter, two convolutional neural network (CNN) models for detection of magnetic anomaly target signals are proposed. One is a 1-D CNN combined with signal feature (SF-CNN1D), and the other is a 2-D CNN based on time–frequency diagrams (TF-CNN2D). To train the models, simulated signals are added to the measured background noise to generate the positive sample set, while the negative sample set is the pure measured noise. Simulation results show that both of the models have satisfying train and test accuracies. A field experiment is conducted to verify the generalization ability of the two CNN models to real data. It is demonstrated that the two CNN models have good detection performances and are some better than the conventional support vector machine (SVM) approach by several percent in colored Gaussian noise scenarios.

The orthogonal basis functions (OBFs) method is a viable approach for magnetic target signal detection under the Gaussian white noise. However, its performance would degrade overtly under the Gaussian color noise with the power spectral density of 1/fα, which is the common environmental magnetic noise. In this letter, based on the linear constrained minimum variance criterion and the generalized likelihood ratio test, respectively, an alternative detection scheme is proposed through the optimization of the basis functions by considering both the signal and noise information. Experiment results using both the simulated data and the measured data show that the present approach can significantly improve the detection performance when compared to the original OBFs method.

In this letter, an appropriate approach is proposed for modeling the electromagnetic (EM) scattering from composite rough surfaces made up of wake due to a submerged body and wind-driven sea waves. The computational fluid dynamics (CFD) method is used to extract the air–seawater surface wake generated by an underwater moving body at different speeds and depths. Then, the wake is superimposed on the randomly rough wind-driven sea surfaces that obey the Pierson–Moskowitz power spectrum. The small slope approximation (SSA) method is adopted to calculate the EM scattering by the composite surfaces. The simulation results are obtained and justified.

Scalar atomic magnetometer (SAM) with high sensitivity is designed to measure the total magnetic field and has been widely applied in aeromagnetic surveys. However, the measured total magnetic field of SAM is influenced by the magnetic disturbances due to the permanent, induced and eddy-current fields of the aircraft and heading errors of SAM. To reduce the magnetic disturbances caused by the aircraft maneuvers, a model was proposed by Tolles and Lawson and has been used in aeromagnetic compensation. However, the heading error of SAM deviates the compensation coefficients estimated by the Tolles-Lawson model. In this letter, we propose a linear model of heading error for multi-cell SAMs and analyze the new aeromagnetic compensation model in which the geomagnetic gradient and heading error are both taken into account. With using a4He magnetometer consisting of three orthogonal cells, the experimental results show that our proposed model has better performances than the previous model in terms of the figure of merit (FOM) and standard deviation (STD).

Numerical modeling of wake produced by an underwater moving object and analysis of its electromagnetic scattering characteristic at microwave bands is developed in this work. The volume of fluid (VOF) model in the computational fluid mechanics is adopted to numerically solve the wake of the underwater moving object. The small slope approximation method is employed to calculate the scattering fields of the wake. Simulation results show that the proposed procedure is viable.

This paper presents a modified method to evaluate the electromagnetic field induced by the wake of a moving slender body in the ocean of finite depth. The electric current induced in the moving seawater by cutting the geomagnetic field will produce additional electromagnetic fields. The additional fields consist of two distinct modes: the oscillating surface wake mode and transient volume wake mode. It is found that, for a given submerged depth, the magnitude of induced magnetic field near the ocean surface is of the order of several nano- Teslas, which is little influenced by the depth of ocean.

A detection procedure for a moving magnetic dipole target by using at least three scalar magnetometers is proposed in this letter. The magnetometers are deployed on the ground or seafloor, while the target is traveling along a straight line at a constant speed above the ground. First, the range from each magnetometer to the closest point of approach (CPA) on the flight line as well as the instant that the target passes by the CPA is estimated by adopting the orthonormal basis function expansion method. Then, the moving parameters, including the moving direction, speed, and elevation, are found through basic geometry relations. Finally, the magnetic moment is determined by solving a linear system by using the least-squares method. Both simulated and field experimental data are employed to demonstrate the proposed scheme.

A new time-reversal imaging method based on the time focusing is presented. The time focusing behavior of the retransmitted time-reversal back-propagating waves from the time-reversal mirror (TRM) at a single target position is extended to multi target positions by using a grouping technique that divides the receiving array as a group of sub-arrays. The number of targets can exceed that of either the transmitters or the receivers. Locally normalized schemes can be employed to favor farer or weaker targets that may be swamped by nearer or stronger targets.

A time domain integral equation (TDIE) method is developed to characterize the pulse responses from a time-evolving ocean-like surface. Pulses echoes are obtained from a series of “frozen” profiles that correspond to the evolving surface at discrete instances. The Doppler shift and bandwidth are examined by using the pulse echoes for different wind speeds and interrogating waveforms. Other than using continuous waves in traditional researches, using pulse waves would give a narrower predicted bandwidth.