In hyperspectral image (HSI) classification, convolutional neural networks (CNNs) and transformer architectures have each contributed to considerable advancements. CNNs possess potent local feature representation skills, whereas transformers excel in learning global features, offering a complementary strength. Nevertheless, both architectures are limited by static receptive fields, which hinder their accuracy in delineating subtle boundary discrepancies. To mitigate the identified limitations, we introduce a novel dual-branch adaptive convolutional transformer (DBACT) network architecture featuring an adaptive multi-head self-attention mechanism. The architecture begins with a triadic parallel stem structure for shallow feature extraction and reduction of the spectral dimension. A global branch with adaptive receptive fields performs high-level global feature extraction. Simultaneously, a local branch with a cross-attention module provides detailed local insights, enriching the global perspective. This methodical integration synergizes the advantages of both branches, capturing representative spatial-spectral features from HSI. Comprehensive evaluation across three benchmark datasets reveals that the DBACT model exhibits superior classification performance compared to leading-edge models.

To find a trajectory with low radar detection probability for stealth aircraft under the assumption of 2D space, performing a rapid turning maneuver is a useful way to reduce the radar detection probability of an aircraft by changing the azimuth angle rapidly to reduce the time of high radar cross-section (RCS) exposure to radar. However, in real flight, not only does the azimuth angle to the radar change rapidly but the elevation angle also changes rapidly, and the change in the radar cross-section is also significant in this process. Based on this premise, this paper established a trajectory planning method based on the sparse A* algorithm in a 3D complex, dynamic environment, called the 3D sparse A* method, based on a log-normal radar model (the 3D-SASLRM method), which considers the RCS statistical uncertainty and the statistical characteristics of the radar signals. Simulations were performed in both simple and complex scenarios. It was concluded that the established 3D-SASLRM method can significantly reduce the radar detection probability. And the essence of reducing under the assumption of 3D space is also to reduce the time of high radar cross-section exposure to radar.

With the continuous improvement and development of armed helicopters, the research on their stealth characteristics has become more and more in depth. In order to obtain the complex effect of stealth characteristics caused by the high-speed rotation of rotor-like components, a dynamic scattering method (DSM) is presented. Rotation speed, azimuth, elevation angle, pitching angle, and rolling angle are studied and discussed in detail. The results show that the electromagnetic scattering characteristics of the main rotor and tail rotor are dynamic and periodic. This period characteristic is related to the rotation speed and attitude angle of the rotor. The radar cross-section (RCS) of the helicopter varies greatly at different observation angles and attitude angles, but the dynamic electromagnetic scattering effect caused by the main rotor and tail rotor cannot be ignored. The presented DSM is effective and efficient for studying the dynamic RCS of the rotor-type parts of a helicopter or the whole machine.

We put forward a new ray tracing method for electromagnetic wave propagation based on piecewise conformal transformations. Our approach is based on our previous research, where conformal transformation was focused on giving an accurate formula for the effective Earth radius. In this article, according to the height interval of atmospheric refractive index data, we approximate the electromagnetic wave propagation path by the combination of circular arcs with different curvature radius and find the effective Earth radius of each arc using conformal transformations. Then, upon using the effective Earth radius, we transform the circular arc path into a straight path, thus simplifying the calculations needed to describe the propagation of electromagnetic waves. We compare our approach to the traditional ray tracing method for several typical atmospheric refractive index profiles and prove that our new method may achieve ray tracing with higher accuracy using fewer path points. We also consider nonuniform atmosphere and irregular terrain and compare the performance of our ray tracing scheme with that of the open-source software PETOOL, which is based on the parabolic equation method. Result proves that the new method may be effectively used for ray tracing in complex environments.

In order to study the radar signature of a tandem helicopter in the X-band, a dynamic analysis approach (DAA) is presented to determine its radar cross-section (RCS) under different influence factors. The basic passage time, rotation speed, observation angle, rotor disk inclination, fuselage attitude angle and Doppler feature are studied and discussed in detail. The results show that the dynamic characteristics of the rotor RCS will bring significant changes to the peak and average values of the helicopter RCS. Within a given observation angle range, a larger elevation angle is undesirable for helicopter stealth. The inclination of the rotor disc will affect the many small peaks and local fluctuations of the helicopter RCS. The positively increased attitude angle will have an undesirable effect on the average RCS and dynamic characteristics of the helicopter. The DAA is feasible and effective for studying the radar cross-section of a tandem helicopter.