Recently, due to the increasing requirements of medical imaging applications and professional annotating images, few-shot learning has gained attention in image semantic segmentation field. To perform with limited number labeled most existing studies use Prototypical Networks (PN) have obtained compelling success. However, these approaches overlook query imagefeatures extracted from proposed representation network, failing preserving spatial connection between support images. In this paper,...

To meet the radar data requirements of target detection technology research and address lack publicly available sea-detecting data, a data-sharing program for is proposed herein. The aim to conduct sea experiments using an X-band solidstate phase-coherent other multi-type radars obtain clutter under different conditions, resolutions, grazing angles. Moreover, marine meteorological hydrological position, track are simultaneously obtained help achieve standardized management radar-measured...

In addressing the issue of weak target detection at sea, most existing feature methods are designed for scenarios with low sea states and small grazing angles. Under high large angles, variations in scattering mechanisms lead to changes characteristics, resulting performance degradation when these applied directly due scene mismatch. To address this, this paper employs four quantitative metrics—mean value, coefficient variation, Bhattacharyya distance, Spearman correlation coefficient—to...

Sea clutter is one of the main limiting factors influencing target detection performance nautical radars. The physical mechanism sea complex with an abundance factors, and non-Gaussian as well non-stationarity behavior significant. Thus, research into property cognition complicated has to be systematic. Based on that concentrates experimental data, this paper reviews summarizes developments in cognition. It properties are most interest for algorithms:amplitude distribution, spectra,...

Reliable and fast detection of maneuvering target in complex background is important for both civilian military applications. It rather difficult due to the motion resulting energy spread time frequency domain. Also, high performance computational efficiency are balance case more pulses. In this paper, we propose a refined processing method radar based on hierarchical detection, utilizing advantages moving (MTD), proposed sparse fractional representation. The adopts two-stage threshold...

In this letter, new models for the spatial correlation of sea clutter texture and intensity are proposed as improved versions current power law or exponential decay model. The have three unknown parameters, thus can be called triparametric models. structure is a weighted sum two components, which describe decaying process coefficient with lags, well periodic behavior due to existence transient coherent structures in clutter. Unknown parameters optimized by nonlinear least square fit method....

Aiming at the technical difficulty of marine radar to detect small targets embedded in sea clutter, this article proposed a three-feature fusion detection method based on singular spectrum analysis. First, considering that number coherent pulses used by scanning mode is usually (64 or less), combines application historical scan data and current frame data, transfers feature extraction from intraframe interframe, extracts three features consist cumulative major value (CMSV), linear degree...

Abstract Motion monitoring systems are often designed and researched to detect the movement of human lower limbs, play an important role in field exoskeleton control. However, current wearable devices can still be improved more convenient or accurate motion recognition. In this work, a comfortable smart gait system was tested. Inertial measurement units (IMUs) flexible membrane compression sensors were implemented, integrated sport pant insoles both feet, respectively. Data acquisition...

For high-resolution radars operating at low grazing angles, sea clutter always exhibits an impulse behavior with heavy tailed statistical property. Presented is a model for spiky based on the generalized central limit theory (CLT). Within compound Gaussian structure, CLT adopted in modeling of Bragg scattering speckle component, while introduced to describe contribution which has potential describing nature underlying mean level because its algebraic (inverse power) tails. Validation results...

Recent attempts to reduce the cost and size of wireless transceivers feature higher levels transistor integration in alternative architectures need for off-chip, high-Q passives used present-day super-heterodyne transceivers. Unfortunately, removal off-chip often comes at increased power consumption circuits preceding including analog-to-digital converter (ADC), which must have dynamic ranges avoid desensitization caused by larger adjacent channel interferers. A selectivity (Q) versus...

Al2O3-SiC-C refractory castables added with β-sialon/Ti(C, N) powders were prepared silica sol and phenolic resin as binding agents at room temperature. The effects of powders, on slag penetration corrosion resistance well oxidation the investigated by X-ray diffraction, scanning electron microscopy energy dispersive spectroscope. Corrosion mechanisms also discussed. results showed that obtained agent a temperature 1450oC presented better resistance. erosion could be improved adding Ti(C,...

The new generation of synthetic aperture radar (SAR) sensors provides us with an opportunity to match multiple high-resolution SAR images. Moreover, the image matching methods have recently gained a lot attention due fact that they can obtain more accurate, better distributed, and reliable matches than stereo methods. In this letter, we present improved multi-image method simultaneously identify from amplitude proposed makes use relationships between pixels in deformed correlation window...

High-speed, low-observable, and maneuvering target detection is always the difficult problem for early warning of space target, ballistic passive radar signal processing. The high-order motion information employed in this paper, such as acceleration jerk. Accordingly, several long-time coherent integration methods are demonstrated illustrated, i.e., Radon-fractional Fourier transform (RFRFT), Radon fractional ambiguity function (RFRAF), Radon-linear canonical (RLCAF), phase...

For high resolution radar at low grazing angles, the existence of strong sea spikes leads to long tails in amplitude distribution clutter. The accurate modeling clutter presence is main contribution this paper. Considering impulsive nature spike events, generalized extreme value (GEV) distribution, which originated from theory (EVT), adopted modeling. Final models are proposed by adding a compound Gaussian component GEV model account for non-spike samples. Validation results with measured...

Due to the difficulty of acquiring underwater target data, recognition small samples acoustic targets has always been a difficult problem in field acoustics. To solve this problem, paper proposes few-shot learning method based on data enhancement and transfer for recognition. This takes two-dimensional time-frequency spectrum as input, uses variety schemes, combined with methods achieve classification. In terms experiments, we used hydrophones collect 3 times different sea areas set...

Herein, a LeNet-based method for wave height inversion and accurate classification of the radar-measured sea clutter data under varying state levels is proposed. Traditionally, empirical formula or spectral integration employed deducing parameters from spectrum. Differently, proposed adopts time-distance two-dimensional data. It combines significant information recorded synchronously during collection to construct training test sets. The neural network trained by random gradient descent...

Traditional radar detection methods heavily rely on the signal-to-clutter ratio (SCR); a variety of feature-based have been proposed, providing new way for and recognition weak targets. Existing determine presence or absence target based whether feature value is within judgment region, generally focusing only distribution features making insufficient use inter-feature chronological information. This paper uses autoregressive (AR) model to predict time sequence echoes in domain takes...

Feature-based target detection methods are predominantly used to determine the presence or absence of targets under sea clutter conditions, but they exhibit a deficiency in making nuanced classification judgments for different categories detected targets. Both ship and marine surface floating small objects display specific characteristics various degrees. Given periodic consistency transient power Doppler centroid bandwidth observed clutter, this paper examines manifestation level...

Intertidal terrain extraction based on satellite remote sensing images is of great military application and civilian value. In this paper, we propose a fusion algorithm SAR optical image to enhance intertidal feature. With the panchromatic from different phases, use an improved shift-invariant discrete wavelet transform obtain higher quality that can not only highlight shape zone, but also characteristics coastline. The real data experiment comparison experiments with other algorithms show...