The integration of Global Positioning Systems (GPS) with Inertial Navigation (INS) has been very actively studied and widely applied for many years. Some sensors artificial intelligence methods have to handle GPS outages in GPS/INS integrated navigation. However, the system using above method still results seriously degraded navigation solutions over long outages. To deal problem, this paper presents a GPS/INS/odometer fuzzy neural network (FNN) land vehicle applications. Provided that...

Previous studies showed that caffeine, diphenhydramine, and carbamazepine were bioconcentrated by mosquito fish (Gambusia holbrooki) from freshwater bodies directly affected reclaimed water. To understand the uptake, depuration, bioconcentration factors (BCFs) under worst-case conditions, authors exposed 84 to water static renewal for 7 d, followed a 14-d depuration phase in clean Characterization of exposure media revealed presence 26 pharmaceuticals, whereas only 5...

Deep reinforcement learning (DRL) has been widely used in many important tasks of communication networks. In order to improve the perception ability DRL on network, some studies have combined graph neural networks (GNNs) with DRL, which use GNNs extract unstructured features network. However, as continue evolve and become increasingly complex, existing GNN-DRL methods still face challenges terms scalability robustness. Moreover, these are inadequate for addressing network security issues....

To detect ships robustly and automatically in monitoring the marine areas, polarimetric synthetic aperture radar imagery is more important. In this letter, three superpixel-level dissimilarity measures are developed to enhance contrast between ship targets sea clutter, which then used construct an automatic detection algorithm. proposed method, multiscale superpixels first generated. Second, measurements a certain superpixel surrounding ones calculated. The transformed from level pixel...

Synthetic aperture radar (SAR) images are especially susceptible to the target aspect angles. For SAR recognition, lack of training data at different angles inevitably deteriorates performance. To solve problem, this letter introduces an angular rotation generative network (ARGN). It is actually attribute-guided transfer learning method, and shared attribute between source domains angle. The angle for each type in domain covers range 0°-360°, while information not complete. Assume that there...

As a current research hotspot, graph convolution networks (GCNs) have provided new opportunities for tree species classification in multi-source remote sensing images. To solve the challenge of limited label information, model was proposed by using semi-supervised fusion method hyperspectral images (HSIs) and multispectral (MSIs). In model, graph-based attribute features pixel-based are fused to deepen correlation improve accuracy. Firstly, employs canonical analysis (CCA) maximize images,...

In order to solve the problem of manual labeling in semi-supervised tree species classification, this paper proposes a pixel-level self-supervised learning model named M-SSL (multisource learning), which takes advantage information plenty multisource remote sensing images and methods. Based on hyperspectral (HSI) multispectral (MSI), features were extracted by combining generative methods with contrastive Two kinds encoders MAAE AAE encoder) MVAE VAE proposed, respectively, set up pretext...

We propose a new framework for camouflaged object detection (COD) named FLCNet, which comprises three modules: an underlying feature mining module (UFM), texture-enhanced (TEM), and neighborhood fusion (NFFM). Existing models overlook the analysis of features, results in extracted low-level texture information that is not prominent enough contains more interference due to slight difference between foreground background object. To address this issue, we created UFM using convolution with...

In this work, we present a camera configuration for acquiring "stereoscopic dark flash" images: simultaneous stereo pair in which one is conventional RGB sensor, but the other sensitive to near-infrared and near-ultraviolet wavelengths instead of red blue. When paired with "dark" flash (i.e., emitting light, no visible light) allows us capture flash/no-flash image at same time, all without disturbing any human subjects or onlookers dazzling flash. We hardware prototype that approximates an...

Magnetic resonance image (MRI) is an important tool to diagnose human diseases effectively. It very for research and clinical application classify the normal abnormal brain MRI images automatically. In this paper, accurate efficient technique proposed extract features of MRIs these into categories. We use two-dimensional multifractal detrended fluctuation analysis (2D MF-DFA) obtain features. These are local generalized Hurst exponents calculated by 2D MF-DFA. regard, values given as...

This article is devoted to experimental study on the control of oblique shock wave around ramp in a low-temperature supersonic flow by means magnetohydrodynamic(MHD) technique. The purpose experiments take advantage MHD interaction weaken strength changing boundary characteristics ramp. Plasma columns are generated pulsed direct current(DC) discharge, magnetic fields Nd-Fe-B rare-earth permanent magnets and waves Lorentz body force effect plasma-induced airflow velocity verified through...

An experimental study of a direct-current, surface arc discharge in Mach 2 cold supersonic airflow is presented. The generated with cylindrical tungsten electrodes flush-mounted on boron-nitride ceramic plate embedded the lower wall test section. In presence airflow, gas breakdown voltage increases from 1.5 kV stationary air to due particle number density augmentation flow. transforms continuous mode pulsed-repetitive mean time interval between pulses about 4.3 ms. For single pulse, occupies...

To detect rotated objects in remote sensing images, researchers have proposed a series of arbitrary-oriented object detection methods, which place multiple anchors with different angles, scales, and aspect ratios on the images. However, major difference between images natural is small probability overlap same category, so anchor-based design can introduce much redundancy during process. In this paper, we convert problem to center point prediction problem, where pre-defined be discarded. By...

Synthetic aperture radar (SAR) image change detection (CD) aims to automatically recognize changes over the same geographic region by comparing prechange and postchange SAR images. However, performance is usually subject several restrictions problems, including absence of labeled samples, inherent multiplicative speckle noise, class imbalance. More importantly, for bitemporal images, changed regions tend present highly variable sizes, irregular shapes, different textures, typically referred...

Satellite image based land cover classification, which falls under the category of semantic segmentation, is critical for many global and environmental applications. Deep learning has been proven to be excellent in segmentation. However, mainstream neural networks formed by connecting high-to-low convolutions series are prone losing information, affects accuracy Besides, it difficult distinguish adjacent classes with similar colors using only RGB information presented satellite images....

Higher standards have been proposed for detection systems since camouflaged objects are not distinct enough, making it possible to ignore the difference between their background and foreground. In this paper, we present a new framework Camouflaged Object Detection (COD) named FSANet, which consists mainly of three operations: spatial detail mining (SDM), cross-scale feature combination (CFC), hierarchical aggregation decoder (HFAD). The simulates three-stage process human visual mechanism...

It has been verified that deep learning methods, including convolutional neural networks (CNNs), graph (GNNs), and transformers, can accurately extract features from hyperspectral images (HSIs). These algorithms perform exceptionally well on HSIs change detection (HSIs-CD). However, the downside of these impressive results is enormous number parameters, FLOPs, GPU memory, training test times required. In this paper, we propose an spectral Kolmogorov-Arnold Network for HSIs-CD (SpectralKAN)....