Small target detection in sea clutter using all-dimensional Hurst exponents of complex time sequence

Small target detection in sea clutter is a challenge problem in radar signal processing community. Based on the analysis of sea clutter Doppler spectrum characteristics, two target detection algorithms are proposed, namely, a Bayesian detection algorithm based on joint Rayleigh distribution model and a feature detection algorithm based on the entropy feature extracted from signal's Doppler spectrum. The detection performances of the proposed algorithms are evaluated based on the data collected by the McMaster IPIX radar at the east coast of Canada.

Low-observable target detection in sea clutter based on the adaptive 3D-IFS algorithm

The polarization information involved in radar echoes could be used to effectively distinguish targets in clutter backgrounds. In this paper, a detection method is proposed based on the characteristics of polarization scattering matrix (PSM) for detecting of weak moving targets in sea clutter. The orientation angle and the amplitude of the maximum return are obtain by the eigenvalue decomposition of the PSM to separate target echo and sea clutter. Meanwhile, the fluctuation of the polarization scattering characteristics between adjacent pulses is also considered with the help of the CP decomposition of the tensor constructed by the PSMs. Based on the feature space which has effectively distinguished target echo and sea clutter, the support vector machine (SVM) is utilized to detect target within sea clutter. Experiments based on real sea clutter data in IPIX database show that the proposed detector could effectively improves the performance for weak targets detection in sea clutter.

The jointed-feature detector for the floating small target in sea clutter is addressed in the paper. For the traditional energy-based detectors, it is difficult to detect the low signal-to-clutter ratio floating small target in time domain due to the affection of sea clutter motion. Therefore, a feature detector in the Fractional Fourier transform (FRFT) domain is proposed. The Hurst exponent and fractal dimension variance are extracted as the features in the jointed-feature detector in FRFT domain. The decision region is determined by convex hull training algorithm on the given false alarm probability. The experimental results of 10 groups of IPIX radar data show that the jointed-feature detector is superior to the compared one, and it provides a new detection scheme for radar target detection.

Doppler frequencies generated by the navigating ship targets may coincide with the first order Doppler spectrum of the sea clutter in high frequency surface wave radar (HFSWR). Therefore, the ship targets fall in this clutter spectrum were usually undetectable by the traditional processing method in HFSWR. In this paper, we introduced a spatial blind filtering method based on oblique projection, which utilized the characteristic difference of ship target and sea clutter returns in spatial directions. Different from the broad beam pattern formed by the conventional digital beamforming technique, the method proposed realized clutter suppressing within the broad beam when the notches are sufficiently dense by blind filtering. The ship target in sea clutter was detected in filed data processing results of HFSWR and it was further verified by the synchronous automatic identification system (AIS) information collected. Those demonstrated that our method can detect the ship targets in sea clutter and this spatial blind filtering technique had a great potential in solving this problem in HFSWR.

In this paper, a novel approach based on the fluctuation analysis for the target detection in sea clutter is proposed. The self-affinity and scaling behaviors of sea clutter is analyzed by using the mean fluctuation. The q-th order normalized slope of the mean fluctuation curve is used as the characteristic parameter to describe the fractal property. The tests on the real data show that the target could be clearly distinguished from the sea clutter background with the proposed approach.

A new knowledge-aided Wald detector is proposed in this paper to deal with the target detection under the inhomogeneous sea clutter. Aim to the scenario of dynamic sea clutter, the proposed detector updates the covariance matrix of background by comprehensive use of the prior information matrix and the covariance matrix. At the same time, the morphological processing is introduced to obtain the final detection result. The experimental results show that the proposed knowledge-aided Wald detector can achieve better detection performance than KA-Wald detector under the inhomogeneous sea clutter.

In the presence of strong sea clutter, target tracking often faces performance degradation in track association and prediction due to high false alarm rates. To address this problem, this paper proposes a radar target detection method based on feedback from dense tracking result regions. The proposed method first applies morphological filtering preprocessing using target tracking results of the current scan cycle. Then, it locates dense clutter regions within the radar scan area by computing connected regions. Based on identified dense clutter regions and a predefined probability of false alarm, the detection threshold for the next scan cycle is adaptively adjusted within these regions. Finally, the updated detection results are fed into the tracker to complete the subsequent target tracking process. Simulation results demonstrate that the proposed method outperforms traditional detect-then-track approaches in terms of target detection and tracking performance under various strong dense sea clutter conditions, while also exhibiting strong robustness.

Small and dim target detection is a critical challenge in hyperspectral remote sensing, particularly in complex, large-scale scenes where spectral variability across diverse land cover types complicates the detection process. In this paper, we propose a novel target reasoning algorithm named Adaptive Global Dense Nested Reasoning Network (AGDNR). This algorithm integrates spatial, spectral, and domain knowledge to enhance the detection accuracy of small and dim targets in large-scale environments and simultaneously enables reasoning about target categories. The proposed method involves three key innovations. Firstly, we develop a high-dimensional, multi-layer nested U-Net that facilitates cross-layer feature transfer, preserving high-level features of small and dim targets throughout the network. Secondly, we present a novel approach for computing physicochemical parameters, which enhances the spectral characteristics of targets while minimizing environmental interference. Thirdly, we construct a geographic knowledge graph that incorporates both target and environmental information, enabling global target reasoning and more effective detection of small targets across large-scale scenes. Experimental results on three challenging datasets show that our method outperforms state-of-the-art approaches in detection accuracy and achieves successful classification of different small targets. Consequently, the proposed method offers a robust solution for the precise detection of hyperspectral small targets in large-scale scenarios.

Because the sea clutter returns acquired for a long duration cannot guarantee the time-stationarity, the estimation performance for the CCM(clutter covariance matrix) is apt to deteriorate even if a sufficient number of secondary data vectors are provided. This paper presents a detection method with clutter-corrupted target returns for a long duration. This is accomplished by dividing the long acquisition period into a number of shorter periods that maintain the time-stationarity so that the CCM for each short duration can be more accurately estimated. To improve detection performance, we show that maximum SINR(signal to interference plus noise ratio) filters designed with the estimated CCMs are applied to each data for a short period and the filter outputs are coherently combined. The improvement in the SINR at the target Doppler bin is theoretically analyzed and verified by simulation.

: Sea clutter has a non-Rayleigh probability density when examined with a radar that has high resolution. This report discusses the properties of non- Rayleigh sea clutter and the detection of targets in non-Rayleigh sea clutter. The spatially varying clutter model is presented and is used to explain the non- Rayleigh nature of sea clutter. The variation of clutter distributions with various radar parameters such as frequency, pulse width, and polarization was determined, and the correlation properties of clutter were established. A way was developed of constructing a realistic realization of the sea surface on a computer, and the surface was used to predict non-Rayleigh clutter densities for various conditions and to indicate some finer points associated with the detection of small targets on the ocean's surface. Finally, various detectors were compared and the problems of pulse-to-pulse and scan-to-scan processing were analyzed.

With the rapid development of UAV and infrared imaging technology, the cost of UAV infrared imaging technology has decreased steadily. Small target detection technology in aerial infrared images has great potential for applications in many fields, especially in the field of traffic surveillance. Because of the low contrast and relatively limited feature information in infrared images compared to visible images, the difficulty involved in small road target detection in infrared aerial images has increased. To solve this problem, this study proposes a feature-enhanced attention and dual-GELAN net (FEADG-net) model. In this network model, the reliability and effectiveness of small target feature extraction is enhanced by a backbone network combined with low-frequency enhancement and a swin transformer. The multi-scale features of the target are fused using a dual-GELAN neck structure, and a detection head with the parameters of the auto-adjusted InnerIoU is constructed to improve the detection accuracy for small infrared targets. The viability of the method was proved using the HIT-UAV dataset and IRTS-AG dataset. According to a comparative experiment, the mAP50 of FEADG-net reached more than 90 percent, which was higher than that of any previous method and it met the real-time requirements. Finally, an ablation experiment was conducted to demonstrate that all three of the modules proposed in the method contributed to the improvement in the detection accuracy. This study not only designs a new algorithm for small road object detection in infrared remote sensing images from UAVs but also provides new ideas for small target detection in remote sensing images for other fields.

In this chapter, we have established a well-defined framework, provided by the likelihood ratio concept and the statistical models developed in Chapters 3 and 6, within which useful detection schemes can be developed in a systematic way. The closely related problem of parameter estimation is also considered: maximum likelihood techniques derived from Bayes' theorem prove to be quite tractable for simple Gaussian clutter models, and can be incorporated into generalised likelihood ratio-based detection methods. Some relatively simple examples of the application of these principles have been discussed in detail, to the point where contact is made with the small target detection strategies discussed in Chapters 12 and 13. The principles demonstrated here can then be applied to other detection scenarios. The estimation of parameters characterising non-Gaussian clutter is more problematic: the maximum likelihood derived equations are now much less easy to solve. Nonetheless, useful estimation methods have been derived that can be applied to gamma, Weibull and K-distributed data. The compound representation of clutter developed in Chapters 3 and 6 plays a central, but rather subtle, role in this work. Small target detection procedures are applied to localised samples of data, to which the relatively simple and tractable Gaussian derived methods can be applied. Consequently, we should expect the methods described in Section 10.7 to be relatively effective even in spiky, non-Gaussian clutter; their performance can then be improved incrementally and relatively straightforwardly, when prior knowledge of the gamma distribution of local power can then be brought to bear, as is described in Section 10.10.

Research on Infrared Dim and Small Target Detection Algorithm Based on Low-Rank Tensor Recovery

Radar detection of small maritime targets can be difficult due to the dynamic motion of the sea. Airborne maritime surveillance platforms are also increasingly required to operate in environments where the sea clutter is strong with persistent breaking waves resembling targets. The focus of this article is to investigate three sparse signal separation formulations using the short time Fourier transform as a dictionary. This approach has been demonstrated as effective in separating both stationary and moving targets from sea clutter but relies on the tuning of different parameters. The first part of this article looks at how to select these parameters which are essential to achieve good separation. Then, a number of practical detection schemes are presented that allow control of the false alarm rate. The algorithm performance is demonstrated using Monte Carlo simulation with synthetic targets injected into the Ingara medium grazing angle sea-clutter data set.