The emerging field of compressed sensing provides sparse reconstruction, which has demonstrated promising results in the areas signal processing and pattern recognition.In this paper, a new approach for synthetic aperture radar (SAR) target classification is proposed based on Bayesian compressive (BCS) with scattering centers features.Scattering features extracted as l 1 -norm problem basis SAR observation physical model, can improve discrimination ability compared original image.Using an...

The inverse synthetic aperture radar (ISAR) imaging for nonuniformly rotating target has always been a challenging task due to the time-varying Doppler parameter, especially in low signal-to-noise ratio (SNR) environment. In this paper, novel ISAR algorithm targets SNR environment based on parameter estimation approach is presented. First, received signal work range bin modeled as multicomponent cubic phase (CPS) after motion compensation. Two approaches, namely coherently integrated...

For inverse synthetic aperture radar (ISAR) imaging of a target with complex motions, the received signal in range bin can be characterized by multicomponent polynomial phase (PPS) after motion compensation. Due to this reason, Fourier transform (FT)-based conventional range-Doppler (RD) ISAR algorithm cannot handle model well, and image obtained will blurred. The synchrosqueezing (SST), which combines continuous wavelet frequency reassignment, is promising tool analyze PPS since it provides...

Change detection is one of the fundamental applications synthetic aperture radar (SAR) images. However, speckle noise presented in SAR images has a negative effect on change detection, leading to frequent false alarms mapping products. In this research, novel two-phase object-based deep learning approach proposed for multi-temporal image detection. Compared with traditional methods, brings two main innovations. One classify all pixels into three categories rather than categories: unchanged...

In this article, a new type of feature, named two-dimensional (2D)-slice Zernike moments, is proposed for synthetic aperture radar (SAR) automatic target recognition (ATR). Target features play an extremely important role in the ATR system. Pixels with different scattering intensities distribute positions SAR images, which represent inherent signatures determined by target’s characteristics, including global structure and local details. To extract these various signatures, we developed...

Automatic target recognition (ATR) is a significant application scenario for Synthetic aperture radar (SAR) image interpretation. In recent years, deep-learning based SAR ATR approaches have made great progress, with the require of large amounts labeled data network training. However, images are scarce, and annotation expensive time-consuming. this paper, novel semi-supervised learning (SSL) framework proposed ATR, which effectively alleviates need samples training, allows excavating...

Polarimetric synthetic aperture radar (PolSAR) image classification plays an important role in various PolSAR application. And many pixel-wise, region-based methods have been proposed for images. However, most of the pixel-wise can not model local spatial relationship pixels due to negative effects speckle noise, and fail figure out regions with similar polarimetric features. Considering that color features provide good visual expression perform well interpretation, this work, based on...

Recently, methods based on deep learning have been applied to target detection using synthetic aperture radar (SAR) images. However, due the SAR imaging mechanism and low signal-clutter-noise-ratio (SCNR), it is still a challenging task perform aircraft imagery. To address this issue, novel method proposed for SCNR images that coherent scattering enhancement fusion attention mechanism. Considering characteristics discrepancy between human-made targets natural background, technique introduced...

In this letter, we propose a synthetic aperture radar (SAR) target configuration recognition algorithm based on the fusion of multifeature low-rank representations (LRRs). First, Gabor, principal component analysis, and wavelet features are extracted for SAR training set test set, respectively. Second, with LRR model, each feature samples is represented by those leading to corresponding coefficient matrix. Then, preliminary prediction labels all sample obtained according coefficients. Third,...

Sparse regression relaxes the difficulties of blind unmixing hyperspectral data thanks to spectral library. Many investigations, however, attach importance global priors such as sparsity and low-rankness. This letter proposes a local-global-based sparse method, called LGSU, by introducing local regularization help boost performance that only considers sparsity. The proposed LGSU first uses superpixel-based technique yield set homogeneous superpixels for guiding purposes. then traditional ℓ...

In response to the inefficiencies and high costs associated with manual buoy inspection, this paper presents design testing of an Autonomous Navigation Unmanned Surface Vehicle (USV) tailored for purpose. The research is structured into three main components: Firstly, hardware framework communication system USV are detailed, incorporating Robot Operating System (ROS) additional nodes meet practical requirements. Furthermore, a tracking utilizing Kernelized Correlation Filter (KCF) algorithm...

Ship detection from synthetic aperture radar (SAR) imagery is crucial for various fields in real-world applications. Numerous deep learning-based detectors have been investigated SAR ship detection, which requires a substantial amount of labeled data training. However, annotation time-consuming and demands specialized expertise, resulting struggling due to lack annotations. With limited data, semi-supervised learning popular approach boosting performance by excavating valuable information...

In this paper, we propose a two-stage multi-task learning representation method for the classification of synthetic aperture radar (SAR) target images. The first stage proposed approach uses multi-features joint sparse learning, modeled as ℓ 2 , 1 -norm regularized problem, to find an effective subset training samples. Then, new dictionary is constructed based on subset. second perform images dictionary, utilizing collaborative representation. algorithm not only exploits discrimination...

An important step in modeling time series is the selection of appropriate model input. Information theoretic concept mutual information provides a general framework to evaluate dependence between potential input and output. A model-free approach, partial measure information, proposed this paper, which utilizes criterion characterize case multiple inputs identifies actual for prediction. This algorithm tested on number synthetic data sets, where attributes were known priori. Results depict...

Recent studies and developments of discrete-time variable structure control (DVSC) theory are summarized in this paper. Some strategies DVSC analyzed. Two problems DVSC-time-delay chattering discussed detail. Finally, the prospective trends suggested.

A new approach to classify synthetic aperture radar (SAR) targets based on high range resolution profiles (HRRPs) is presented. Features from each of the target HRRPs are extracted via nonnegative matrix factorization (NMF) algorithm in time-frequency domain represented by adaptive Gaussian representation (AGR). Firstly, SAR images have been converted into HRRPs. And for obtained using AGR. Secondly, feature vectors utilizing NMF. Finally, hidden Markov models (HMMs) employed characterize...

This paper first describes the research background of inverted pendulum, then derives mathematical model planar double pendulum system by use Lagrange modeling method. After partial linearization, we can arrive at pendulum's equation state space, which is a controllable, observable, but absolute unstable after analyzing. Using pole-placement method, variable structure controller was designed for stabilization and robust control pendulum. In order to reduce chattering controller, new...

The paper deals with a novel method, self-adjustable fuzzy sliding mode control, of squirrel cage induction motor drive systems voltage source inverter. According to the principle logic control and theory this new is presented mathematical description for derived. used achieve robust performance against parameter variations external disturbances. In order reduce or overcome system chattering owing law employed draw state variables into pre-specified bounds surface, stator designed be...