Improved Small Target Detection Method for SAR Image Based on YOLOv7

In order to solve the problems that the current synthetic aperture radar(SAR)image target detection method cannot adapt to targets of different sizes,and the complex image background leads to low detection accuracy,an improved SAR image small target detection method based on YOLOv7 was proposed in this study.The proposed method improved the feature extraction network by using Switchable Around Convolution(SAConv)in the backbone network to help the model capture target information at different scales,thus improving the feature extraction ability for small targets.Based on the attention mechanism,the DyHead module was embedded in the target detection head to reduce the impact of complex background,and better focus on the small targets.In addition,the NWD loss function was introduced and combined with CIoU loss.Compared to the CIoU loss function typically used in YOLOv7,the NWD loss function pays more attention to the processing of small targets,so as to further improve the detection ability of small targets.The experimental results on the HRSID dataset indicate that the proposed method achieved mAP@0.5 and mAP@0.95 scores of 93.5%and 71.5%,respectively.Compared to the baseline model,this represents an increase of 7.2%and 7.6%,respectively.The proposed method can effectively complete the task of SAR image small target detection.

Novel method for extraction of ship target with overlaps in SAR image via EM algorithm

The quality of synthetic aperture radar(SAR)image degrades in the case of multiple imaging projection planes(IPPs)and multiple overlapping ship targets,and then the performance of target classification and recognition can be influenced.For addressing this issue,a method for extracting ship targets with overlaps via the expectation maximization(EM)algorithm is pro-posed.First,the scatterers of ship targets are obtained via the target detection technique.Then,the EM algorithm is applied to extract the scatterers of a single ship target with a single IPP.Afterwards,a novel image amplitude estimation approach is pro-posed,with which the radar image of a single target with a sin-gle IPP can be generated.The proposed method can accom-plish IPP selection and targets separation in the image domain,which can improve the image quality and reserve the target information most possibly.Results of simulated and real mea-sured data demonstrate the effectiveness of the proposed method.

Method of moving target detection based on sub-image cancellation for single-antenna airborne synthetic aperture radar

The method of moving target detection based on subimage cancellation for single-antenna airborne SAR is presented. First the subimage is obtained through frequency processing is pointed out. The imaging difference of a stationary objects and moving object in the subimage based on the frequency division is analyzed from the fundamental principle. Then the developed method combines the shear averaging algorithm to focus on the moving target in the subimage, after the clutter suppression and the focusing position in each subimage is obtained. Next the observation model and the relative movement of the moving targets between the subimages estimate the moving targets. The theoretical analysis and simulation results demonstrate that the method is effective and can not only detect the moving targets, but also estimate their motion parameters precisely.

A Hybrid Features Based Detection Method for Inshore Ship Targets in SAR Imagery

Convolutional Neural Networks(CNNs)have recently attracted much attention in the ship detection from Synthetic Aperture Radar(SAR)images.However,compared with optical images,SAR ones are hard to understand.Moreover,due to the high similarity between the man-made targets near shore and inshore ships,the classical methods are unable to achieve effective detection of inshore ships.To mitigate the influence of onshore ship-like objects,this paper proposes an inshore ship detection method in SAR images by using hybrid features.Firstly,the sea-land segmentation is applied in the pre-processing to exclude obvious land regions from SAR images.Then,a CNN model is designed to extract deep features for identifying potential ship targets in both inshore and offshore water.On this basis,the high-energy point number of amplitude spectrum is further introduced as an important and delicate feature to suppress false alarms left.Finally,to verify the effectiveness of the proposed method,numerical and comparative studies are carried out in experiments on Sentinel-1 SAR images.

A novel oil spill detection method from synthetic aperture radar imageries via a bidimensional empirical mode decomposition

Oil spills pose a major threat to ocean ecosystems and their health. Synthetic aperture radar（SAR） sensors can detect oil spills on the sea surface. These oil spills appear as dark spots in SAR images. However, dark formations can be caused by a number of phenomena. It is aimed to distinguishing oil spills or look-alike objects. A novel method based on a bidimensional empirical mode decomposition is proposed. The selected dark formations are first decomposed into several bidimensional intrinsic mode functions and the residue. Subsequently, 64 dimension feature sets are calculated using the Hilbert spectral analysis and five new features are extracted with a relief algorithm. Mahalanobis distances are then used for classification. Three data sets containing oil spills or look-alikes are used to test the accuracy rate of the method. The accuracy rate is more than 90%. The experimental results demonstrate that the novel method can detect oil spills validly and accurately.

A Mo LC+Mo M-based G^0 distribution parameter estimation method with application to synthetic aperture radar target detection

The accuracy of background clutter model is a key factor which determines the performance of a constant false alarm rate(CFAR) target detection method. G0 distribution is one of the optimal statistic models in the synthetic aperture radar(SAR) image background clutter modeling and can accurately model various complex background clutters in the SAR images. But the application of the distribution is greatly limited by its disadvantages that the parameter estimation is complex and the local detection threshold is difficult to be obtained. In order to solve the above-mentioned problems, an synthetic aperture radar CFAR target detection method using the logarithmic cumulant(Mo LC) + method of moment(Mo M)-based G0 distribution clutter model is proposed. In the method, G0 distribution is used for modeling the background clutters, a new Mo LC+Mo M-based parameter estimation method coupled with a fast iterative algorithm is used for estimating the parameters of G0 distribution and an exquisite dichotomy method is used for obtaining the local detection threshold of CFAR detection, which greatly improves the computational efficiency, detection performance and environmental adaptability of CFAR detection. Experimental results show that the proposed SAR CFAR target detection method has good target detection performance in various complex background clutter environments.

Power of SAR Imagery and Machine Learning in Monitoring Ulva prolifera:A Case Study of Sentinel-1 and Random Forest

Automatically detecting Ulva prolifera(U.prolifera)in rainy and cloudy weather using remote sensing imagery has been a long-standing problem.Here,we address this challenge by combining high-resolution Synthetic Aperture Radar(SAR)imagery with the machine learning,and detect the U.prolifera of the South Yellow Sea of China(SYS)in 2021.The findings indicate that the Random Forest model can accurately and robustly detect U.prolifera,even in the presence of complex ocean backgrounds and speckle noise.Visual inspection confirmed that the method successfully identified the majority of pixels containing U.prolifera without misidentify-ing noise pixels or seawater pixels as U.prolifera.Additionally,the method demonstrated consistent performance across different im-ages,with an average Area Under Curve(AUC)of 0.930(+0.028).The analysis yielded an overall accuracy of over 96%,with an aver-age Kappa coefficient of 0.941(+0.038).Compared to the traditional thresholding method,Random Forest model has a lower estima-tion error of 14.81%.Practical application indicates that this method can be used in the detection of unprecedented U.prolifera in 2021 to derive continuous spatiotemporal changes.This study provides a potential new method to detect U.prolifera and enhances our under-standing of macroalgal outbreaks in the marine environment.

SAR target detection based on the optimal fractional Gabor spectrum feature

In this paper,an algorithm based on a fractional time-frequency spectrum feature is proposed to improve the accuracy of synthetic aperture radar(SAR)target detection.By extending the fractional Gabor transform(FrGT)into two dimensions,the fractional time-frequency spectrum feature of an image can be obtained.In the achievement process,we search for the optimal order and design the optimal window function to accomplish the two-dimensional optimal FrGT.Finally,the energy attenuation gradient(EAG)feature of the optimal time-frequency spectrum is extracted for high-frequency detection.The simulation results show the proposed algorithm has a good performance in SAR target detection and lays the foundation for recognition.

Arbitrary-oriented target detection in large scene sar images

Target detection in the field of synthetic aperture radar(SAR) has attracted considerable attention of researchers in national defense technology worldwide,owing to its unique advantages like high resolution and large scene image acquisition capabilities of SAR.However,due to strong speckle noise and low signal-to-noise ratio,it is difficult to extract representative features of target from SAR images,which greatly inhibits the effectiveness of traditional methods.In order to address the above problems,a framework called contextual rotation region-based convolutional neural network(RCNN) with multilayer fusion is proposed in this paper.Specifically,aimed to enable RCNN to perform target detection in large scene SAR images efficiently,maximum sliding strategy is applied to crop the large scene image into a series of sub-images before RCNN.Instead of using the highest-layer output for proposal generation and target detection,fusion feature maps with high resolution and rich semantic information are constructed by multilayer fusion strategy.Then,we put forwards rotation anchors to predict the minimum circumscribed rectangle of targets to reduce redundant detection region.Furthermore,shadow areas serve as contextual features to provide extraneous information for the detector identify and locate targets accurately.Experimental results on the simulated large scene SAR image dataset show that the proposed method achieves a satisfactory performance in large scene SAR target detection.

Micro-motion effect in inverse synthetic aperture radar imaging of ballistic mid-course targets

For ballistic mid-course targets,in addition to constant orbital motion,the target or any structure on the target undergoes micro-motion dynamics,such as spin,precession and tumbling.The micro-motion characteristics of the ballistic mid-course targets were discussed.The target motion model and inverse synthetic aperture radar(ISAR) imaging model for this kind of targets were built.Then,the influence of micro-motion on ISAR imaging based on the established imaging model was presented.The computer simulation to get mid-course target echoes from static darkroom electromagnetic scattering data based on the established target motion model was realized.The imaging results of computer simulation show the validity of ISAR imaging analysis for micro-motion targets.

A NOVEL SVM FOR GROUND PENETRATING SYNTHETIC APERTURE RADAR LANDMINE DETECTION

The use of vehicle- or air-borne Ground Penetrating Synthetic Aperture Radar (GPSAR) to quickly detect landmines over large areas is becoming a trend. However, producing too many false alarms in GPSAR landmine detection is a major challenge in practical applications of GPSAR. Support Vector Machine (SVM), employing structural risk minimization theory, does not need large amounts of training data, which makes it suitable for solving the landmine detection problem. In this paper, a novel SVM with a hypersphere instead of a hyperplane classification boundary is proposed for landmine detection in GPSAR. The HyperSphere-SVM (HS-SVM) can be trained with both landmine and clutter data, or with landmine data only, which are called the two-class HS-SVM and the one-class HS-SVM, respectively. The HS-SVM has better generalization capability than the traditional HyperPlane-SVM (HP-SVM) with respect to varying operating conditions. Quantitative comparisons have been made using real data collected with the rail-GPSAR landmine detection system, which show that both the two-class and the one-class HS-SVMs have better detection performance than the HP-SVM.

Target detection and recognition in SAR imagery based on KFDA

Current research on target detection and recognition from synthetic aperture radar （SAR） images is usually carried out separately. It is difficult to verify the ability of a target recognition algorithm for adapting to changes in the environment. To realize the whole process of SAR automatic target recognition （ATR）, es- pecially for the detection and recognition of vehicles, an algorithm based on kernel fisher discdminant analysis （KFDA） is proposed. First, in order to make a better description of the difference be- tween the background and the target, KFDA is extended to the detection part. Image samples are obtained with a dual-window approach and features of the inner and outer window samples are extracted by using KFDA. The difference between the features of inner and outer window samples is compared with a threshold to determine whether a vehicle exists. Second, for the target area, we propose an improved KFDA-IMED （image Euclidean distance） combined with a support vector machine （SVM） to recognize the vehicles. Experimental results validate the performance of our method. On the detection task, our proposed method obtains not only a high detection rate but also a low false alarm rate without using any prior information. For the recognition task, our method overcomes the SAR image aspect angle sensitivity, reduces the requirements for image preprocessing and improves the recogni- tion rate.

MAXIMUM A POSTERIORI-BASED AUTOMATIC TARGET DETECTION IN SAR IMAGES

The paper presents an algorithm of automatic target detection in Synthetic Aperture Radar(SAR) images based on Maximum A Posteriori(MAP). The algorithm is divided into three steps. First, it employs Gaussian mixture distribution to approximate and estimate multi-modal histogram of SAR image. Then, based on the principle of MAP, when a priori probability is both unknown and learned respectively, the sample pixels are classified into different classes c = {target,shadow, background}. Last, it compares the results of two different target detections. Simulation results preferably indicate that the presented algorithm is fast and robust, with the learned a priori probability, an approach to target detection is reliable and promising.

SAR Moving Target Detection and Imaging Based on WVH Transform

Aim To propose a generalized and closed representation of the Wigner Ville Hough transform(WVHT), for the moving target detection and imaging in the design of synthetic aperture radar(SAR). Methods Based on the line integral, the WVH transform was derived by combining the Wigner Ville distribution (WVD) and the Hough transform (HT) together. The new transform was then verified with computer by the simulated SAR echoes. Results and Conclusion The correctness and the validity of the WVH transform were proved by the computer simulation. Compared with the conventional WVD HT method, the new approach based on the WVHT can simplify the processing procedure, it can translate the chirp echoes of multi targets of SAR from the time domain into the parameter space directly, while suppressing the cross terms of WVD and estimating the motion coefficients for the final imaging. It is obvious that the WVH transform can be also used in other cases for the chirp signal detection.

微波视觉与SAR图像智能解译

高分辨率雷达成像技术和人工智能、大数据技术的快速发展,有力促进了雷达图像智能解译技术的进步。由于雷达传感器本身的特殊性和电磁散射成像物理的复杂性,雷达图像的解译缺乏光学图像的直观性,准确迅速识别分类的需求对雷达图像解译提出了迫切的挑战。在借鉴人脑光视觉感知机理和计算机视觉图像处理相关技术基础上,进一步融合电磁散射物理规律及其雷达成像机理,我们提出发展微波域雷达图像解译的“微波视觉”的新交叉领域研究。该文介绍微波视觉的概念与内涵,提出微波视觉认知模型,阐述其基础理论问题与技术路线,最后介绍了作者团队在相关问题上的初步研究进展。

基于局部显著特征聚焦学习的SAR舰船智能检测

合成孔径雷达(Synthetic Aperture Radar,SAR)图像的船舶目标检测,因其广泛的应用前景而备受关注。近年来,基于深度学习的SAR图像船舶目标检测在多种场景中表现出较好性能。然而,由于SAR独特的成像机制,舰船目标通常与背景环境具有相似的散射特性使得实际的船舶目标难以辨识,且船舶目标尺度较小,导致准确检测船舶目标具有挑战性。为了缓解这一问题,本文提出了一种基于局部显著特征聚焦学习的SAR舰船检测方法。首先,设计了双重注意力模块,通过对通道级和空间级的特征进行双重注意力加权,以充分地探索船舰目标的关键语义特征,从而提升模型的深度提取能力。随后,为了进一步提升模型对船舶目标特征的表征能力,设计了平衡特征金字塔网络模块,通过对舰船目标的多尺度特征进行缩放、增强和聚合处理,以实现多尺度特征间的语义和空间信息均衡分布。最后,在SAR舰船检测数据集(SAR Ship Detection Dataset,SSDD)上进行了广泛的实验分析,实验结果一致性地证明了所提方法在提升SAR图像舰船目标检测准确性方面的有效性。

DEVELOPMENT OF MOVING TARGET DETECTION AND IMAGING BY AIRBORNE SAR

The detection and ima ging of moving targets based on airborne synthetic aperture radar (SAR) is a cru cial technique for the modern radar. Firstly, the mathematical model of SAR ech o signal which comes from moving targets is constructed. Based on this model, th e features of moving target imaging are introduced and the effects of target mov ement to SAR imaging are analyzed. Then the development and the status of this t echnique are reviewed in detail. Finally, some frontiers of this field are point ed out.

多域特征引导的无监督SAR图像舰船检测方法

如何在合成孔径雷达(SAR)图像标注样本有限的条件下,提升舰船检测性能一直是SAR图像处理中的热点问题。本文提出一种多域特征引导的无监督域适应方法,将知识从有标注的源域(光学图像)转移到未标注的目标域(SAR图像),降低对标记SAR图像数据依赖。同时,设计了频域转换模块、注意力区域增强模块和自适应权重模块来缩小光学、SAR图像域之间的域差距,提高源域与目标域特征对齐效率,增强网络在挑战性样本下的特征迁移能力。在公开发布的数据集上进行了大量实验。结果表明:所提的模块较基础模型AP50提升10%,总体性能优于其他先进的方法。

考虑速度聚束效应的SAR海浪成像仿真方法

针对合成孔径雷达(SAR)海浪成像仿真技术对于速度聚束效应的考虑不充分,仿真数据无法准确反映实际SAR数据的问题,建立一种充分考虑速度聚束效应的SAR海浪成像仿真方法并进行了实验验证。计算了SAR图像强度的概率密度分布的仿真结果与理论结果之间的均方误差(MSE),在风速分别为5、10和15 m/s时,考虑速度聚束效应前、后比对结果的MSE分别为0.1012、0.1576、0.0556与0.0179、0.0314、0.0088。仿真结果表明,所提SAR海浪成像仿真方法有效提高了仿真数据的准确性,对SAR海浪成像的应用具有一定实用价值。

语义增强与高阶强交互的SAR图像舰船检测

合成孔径雷达(synthetic aperture radar,SAR)图像背景信息复杂、舰船目标边缘模糊,且多为容易丢失的小尺度舰船目标。针对上述问题,提出语义增强与高阶强交互的SAR图像舰船检测。该方法利用部分卷积与非对称卷积构建部分非对称卷积聚合网络,在减少计算复杂度、轻量化主干网络的同时,更好地捕捉多尺度舰船特征,同时在上采样部分引入双层路由注意力,增强对图像上下文信息的利用。另外,通过递归的方式进行特征提取,可以较好解决区域内信息交互的问题,实现不同级别特征之间的高阶交互建模,提升模型检测能力。在公开的HRSID遥感数据集上进行实验的结果表明,该方法的检测精度达到91.23%,相比原模型提升5.13%,准确率与召回率分别提升2.41%和7.16%,与主流算法相比具有较好的检测效果。