This letter addresses the problem of adaptive coherent detection of maritime high-resolution radar range-spread targets in correlated heavy-tailed sea clutter. We first model radar sea clutter by the compound Gaussian model with lognormal texture and unknown speckle covariance matrix. The lognormal-distributed texture can capture the tail-level of sea clutter, and the speckle covariance matrix contains the pulse-to-pulse correlation of sea clutter. Then, based on the two-step generalized likelihood ratio test and the maximum likelihood or a posteriori estimation of unknown parameters, an adaptive coherent generalised likelihood ratio test with lognormal texture detector is proposed to detect radar range-spread targets. The proposed detector has the ability to be adaptive to clutter power mean, non-Gaussianity and pulse-to-pulse correlation. The performance evaluation experiments on simulated and measured data show that the proposed detector outperforms conventional adaptive detectors. More specifically, the detection results on measured data indicate that when the number of target range cells is 3 and the probability of detection reaches 0.8, the proposed detector has a signal-to-clutter ratio gain of about 1 dB over its competitors.

This letter deals with the knowledge-aided adaptive detection problem of radar targets in non-homogenous correlated compound Gaussian sea clutter. The lognormal distribution is used as the prior distribution of the clutter texture to match the non-Gaussianity of sea clutter. In addition, in order to ensure the estimation accuracy of the covariance matrix structure of sea clutter, a convex combination estimator (CCE) is proposed by jointly exploiting the prior information and the current secondary data. Then, a knowledge-aided adaptive detector is designed on the basis of the complex parameter Gradient test and the CCE. Numerical experiments verify the effectiveness of the proposed CCE and adaptive detector in comparison with their counterparts, respectively.

This article presents a classification method to classify different marine floating small targets, which can realize effective classification of different targets in strong clutter background. The design of proposed classification method is primarily based on block-whitened time–frequency spectrogram and pre-trained convolution neural network (CNN). Block-whitening clutter suppression is used to process target echoes. By converting a strong clutter background to an approximately noisy background, the effect of strong clutter on classification is reduced. Then, the time–frequency spectrogram of targets is extracted from the block-whitened target echoes, which converts a signal in time domain into a time–frequency spectrogram with more information. In addition, the block-whitened time–frequency spectrograms are input to a pre-trained CNN for feature extraction and classification training. By exploiting pre-training procedure, the proposed method can effectively classify different marine floating small targets and solve the problem of limited target samples in practical applications. Finally, a dataset of three kinds of measured maritime radar targets is constructed to verify the effectiveness of proposed method. Experimental results show that compared with competitors, the pre-trained CNN with block-whitened time–frequency spectrograms can achieve higher performance on the measured dataset.

In this paper, we consider the detection problem of marine radar targets embedded in correlated non-Gaussian sea clutter, which is modelled by a compound Gaussian model with lognormal texture (CG-LN) and unknown covariance matrices. In order to reduce the dependence of detectors on training data, the original radar data are transformed via exploiting the persymmetric structure of clutter covariance matrix. Then, four complex parameter suboptimum tests, which are the Rao, Wald, gradient, and Durbin tests, are utilized to design the adaptive persymmetric coherent detectors for radar targets. It is shown that the Gradient test and the Durbin test coincide with the Rao test for the problem of radar target detection in CG-LN sea clutter. We prove that the proposed persymmetric Rao detector with lognormal texture (PRAO-LND) and persymmetric WALD detector with lognormal texture (PWALD-LND) can ensure the constant false alarm rate with respect to the clutter power mean and the clutter speckle covariance matrix. Experimental results on simulated and measured radar data show that the proposed PRAO-LND performs better than its competitors, and is robust to the mismatched signals. Moreover, the proposed PWALD-LND has asymptotic performance with the proposed PRAO-LND, when the non-Gaussianity of sea clutter is weakened, and has good selectivity with signal mismatch.

This paper deals with the detection problem of a moving point-like target in correlated non-Gaussian sea clutter, which is modelled by a compound Gaussian model with a lognormal-distributed texture and an unknown covariance matrix. In order to improve the detection performance for radar targets in sample-starved environments where the number of secondary data is limited, the persymmetric structure is exploited to transform the original radar data. Based on the two-step generalized likelihood ratio test (GLRT) and its maximum a posterior version, we propose two adaptive persymmetric coherent detectors for radar target detection. Theoretical and experimental confirmations are provided to show that the proposed detectors guarantee the constant false alarm rate property with respect to the clutter covariance matrix structure and the clutter power mean. Experimental results on simulated and measured radar data demonstrate that two proposed detectors perform better than traditional ones, especially when the number of secondary data is small.

In this paper, we consider the problem of robustly and accurately estimating the shape parameters ofK-distributed sea clutter in the maritime radar industry. Outliers formed by non-sea-surface echoes have a significant negative impact on the estimation accuracy. To improve the estimation performance, we first propose a bipercentiles feedforward neural network for the shape parameter$\eta$(BP-FFNN-$\eta$), which utilizes a ratio of two percentiles and a two-hidden-layer feedforward neural network. The BP-FFNN-$\eta$can learn the mathematical relationship between the shape parameter and the ratio of two percentiles, and can work in environments where the number of outliers is approximately known. Moreover, to solve the case where the number of outliers is not known due to dynamic changes in the environment, we also design another neural network (referred to as MBP-FFNN-$\eta$), which consists of multiple BP-FFNNs-$\eta$and a multi-class classification network. The MBP-FFNN-$\eta$can perceive the change in the proportion of outliers, so an accurate estimate can be obtained from an unaffected BP-FFNN-$\eta$. Finally, training and test data are constructed to train and evaluate the proposed methods, respectively. Experimental results demonstrate that the BP-FFNN-$\eta$performs better than traditional moments-based estimators, and has almost the same performance as the tri-percentile estimator. Compared with the tri-percentile estimator, the BP-FFNN-$\eta$avoids table lookups, and produces a continuous estimate. The MBP-FFNN-$\eta$can achieve more than 97% overall classification accuracy on simulated and measured data, and thus an accurate estimate of the shape parameter can be obtained when the number of outliers varies.

We present a spatial–spectral decoupling interaction network for multispectral imagery change detection, which can exploit the underlying information of the multispectral imagery adequately through simultaneously considering the discriminative attribute of each pixel and robust spatial structure of the corresponding patch. Specifically, a 1-D convolutional neural network (1D-CNN) is applied to the spectral vector of each pixel to extract its discriminative feature, while a 2D-CNN is applied to the patch centering on the corresponding pixel to explore the spatial structure information. In addition, an interaction mechanism is incorporated into the feature fusion module to enhance the spatial–spectral consistency.

We consider the detection problem of maritime radar targets in the training-sample-starved and non-Gaussian sea clutter environment. The performance of conventional detectors for radar targets is seriously degraded due to both the starvation of training samples for estimating the clutter covariance matrix and the non-Gaussianity of sea clutter. In this letter, we adopt the inverse Gaussian distribution and the inverse complex Wishart distribution to model the texture and speckle covariance matrix of sea clutter, respectively. Then an adaptive Bayesian detector is developed based on the two-step generalized likelihood ratio test and the maximum posterior estimates of clutter parameters. Finally, the experimental results on simulated and measured data demonstrate the performance superiority of the proposed detector over its competitors, especially when the training samples are starved.

This article deals with the problem of detecting a marine target with coherent radars in a correlated heavy-tailed sea clutter background. The heavy-tailed sea clutter is modeled by a compound-Gaussian model, and the clutter texture is characterized by the lognormal distribution with a new parameterization form. We develop an adaptive coherent detector on the basis of the two-step generalized likelihood ratio test. The proposed detector can achieve adaptation to sea clutter characteristics by using the maximum a posterior estimate of the clutter texture, the constrained approximate maximum likelihood estimator of the speckle covariance matrix, and the proposed negative- and positive-fractional moment estimate of amplitude parameters of sea clutter. Remarkably, the proposed detector inherently ensures a constant false alarm rate with respect to the clutter power mean and the speckle covariance matrix. Finally, numerical experiments using simulated data and real radar data demonstrate that the proposed estimator and adaptive coherent detector outperform their respective competitors.

Non-Gaussian sea clutter causes conventional detectors designed in Gaussian clutter to suffer detection performance degradation, and some nuisance parameters cause the uniformly most powerful test to be unavailable. To improve the detection performance of maritime radar targets, we investigate the design of adaptive detectors in correlated non-Gaussian sea clutter via using suboptimal tests. The non-Gaussian sea clutter is modelled as a product of lognormal-distributed texture and complex Gaussian speckle. Two adaptive radar target detectors are developed by using the suboptimal two-step Wald and Rao tests. Specifically, a non-adaptive detector is derived by the Wald or Rao test when the clutter texture and speckle covariance matrix are assumed to be known in the first step; then the clutter parameters known in the first step are estimated, and the true parameters of the detector obtained in the first step are replaced with the estimated values. Theoretical proof and experimental verification indicate that the two proposed detectors have the constant false alarm property with regard to the clutter speckle covariance matrix and the clutter average power. Numerical results on simulated and measured radar data show that the proposed Rao-based detector outperforms its competitors, and has the stronger robustness to the signal mismatch compared to the proposed Wald-based detector.

This article addresses the detection problem of radar targets embedded in nonhomogeneous and non-Gaussian sea clutter. Nonhomogeneity leads to insufficiency of secondary data for estimating the clutter speckle covariance matrix, and non-Gaussianity causes sea clutter to become spiky. In this article, the persymmetry of the clutter covariance matrix is adopted to alleviate the requirement of secondary data, and the prior distribution of clutter texture is exploited to tackle the clutter non-Gaussianity. Based on such clutter knowledge, three adaptive detectors are proposed according to the principles of the generalized likelihood ratio test, the Wald test, and the Rao test. It is proven that three detectors ensure constant false alarm rate (CFAR) properties with respect to both the clutter speckle covariance matrix and the clutter power mean. Simulation experiments show that three detectors outperform their competitors.

In this letter, we focus on the statistical modeling of sea clutter amplitudes. Due to its non-Gaussian nature, the existing statistical models are sometimes difficult to represent well the heavy-tailed portion of amplitude distribution. To address this problem, we propose a compound Gaussian (CG) model with a generalized inverse Gaussian (GIG) texture to describe sea clutter amplitudes. In this regard, the probability density function and the cumulative distribution function of the clutter amplitudes for the proposed model are derived. Moreover, we provide an approach to estimate the unknown parameters of the proposed CG-GIG distribution. The experimental results indicate that the CG-GIG distribution is more suitable to describe the amplitudes of non-Gaussian sea clutter than its competitors.