A Fast Threshold Neural Network for Ship Detection in Large-Scene SAR Images

Cui, J. X.; Jia, Hecheng; Wang, Haipeng; Xu, Feng

doi:10.1109/jstars.2022.3192455

Cited by 23 publications

(12 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Target proposal generation can be achieved by network modules as well, i.e., networks can be used to jointly locate and identify the targets. Representative methods include SSD, Faster-RCNN, YOLO, and other deep network-based detectors [18,19]. With the development of deep learning techniques, novel network modules and learning paradigms have been introduced in the field of SAR as well, such as attention mechanism [20,21], transfer learning [22], semi-supervised learning [23], etc.…”

Section: Deep Learning-based Methodsmentioning

confidence: 99%

Lightweight CFARNets for Landmine Detection in Ultrawideband SAR

Zhang,

Song,

Jin

2023

Remote Sensing

View full text Add to dashboard Cite

The high-resolution image obtained by ultrawideband synthetic aperture radar (UWB SAR) includes rich features such as shape and scattering features, which can be utilized for landmine discrimination and detection. Due to the high performance and automatic feature learning ability, deep network-based detection methods have been widely employed in SAR target detection. However, existing deep networks do not consider the target characteristics in SAR images, and their structures are too complicated. Therefore, lightweight deep networks with efficient and interpretable blocks are essential. This work investigates how to utilize the SAR characteristics to design a lightweight deep network. The widely employed constant false alarm rates (CFAR) detector is used as a prototype and transformed into trainable multiple-feature network filters. Based on CFAR filters, we propose a new class of networks called CFARNets which can serve as an alternative to convolutional neural networks (CNNs). Furthermore, a two-stage detection method based on CFARNets is proposed. Compared to prevailing CNNs, the complexity and number of parameters of CFARNets are significantly reduced. The features extracted by CFARNets are interpretable as CFAR filters have definite physical significance. Experimental results show that the proposed CFARNets have comparable detection performance compared to other real-time state-of-the-art detectors but with faster inference speed.

show abstract

Section: Deep Learning-based Methodsmentioning

confidence: 99%

Lightweight CFARNets for Landmine Detection in Ultrawideband SAR

Zhang,

Song,

Jin

2023

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…For the width and height of negative sample bounding boxes, we want the variation to be between 50% and 100% of the width and height of the real samples, while ensuring that the smallest bounding box of the negative samples is not less than 10% of the real sample size. We want the noisy positive and negative samples to have no significant variation from the real samples, which will help the network better restore the original coordinates of the instance boxes [ 15 ]. For positive samples, the decoder learns the corresponding real sample boxes, while negative samples represent “non-existent objects”.…”

Section: Algorithm Frameworkmentioning

confidence: 99%

“…These improvements mainly include using stronger backbone networks [ 8 , 9 ], setting up multi-scale FPN layers [ 10 , 11 ], and designing loss functions more suitable for SAR tasks [ 12 , 13 ]. Meanwhile, real-time SAR target detection schemes [ 14 , 15 , 16 ] are also gradually developing, providing references for the practical application of SAR detection and recognition. Moreover, the transformer [ 17 ] architecture has demonstrated remarkable efficacy in optical image recognition by leveraging the global modeling capabilities of its self-attention mechanism.…”

Section: Introductionmentioning

confidence: 99%

CCDN-DETR: A Detection Transformer Based on Constrained Contrast Denoising for Multi-Class Synthetic Aperture Radar Object Detection

Zhang,

Zheng,

et al. 2024

Sensors

View full text Add to dashboard Cite

The effectiveness of the SAR object detection technique based on Convolutional Neural Networks (CNNs) has been widely proven, and it is increasingly used in the recognition of ship targets. Recently, efforts have been made to integrate transformer structures into SAR detectors to achieve improved target localization. However, existing methods rarely design the transformer itself as a detector, failing to fully leverage the long-range modeling advantages of self-attention. Furthermore, there has been limited research into multi-class SAR target detection. To address these limitations, this study proposes a SAR detector named CCDN-DETR, which builds upon the framework of the detection transformer (DETR). To adapt to the multiscale characteristics of SAR data, cross-scale encoders were introduced to facilitate comprehensive information modeling and fusion across different scales. Simultaneously, we optimized the query selection scheme for the input decoder layers, employing IOU loss to assist in initializing object queries more effectively. Additionally, we introduced constrained contrastive denoising training at the decoder layers to enhance the model’s convergence speed and improve the detection of different categories of SAR targets. In the benchmark evaluation on a joint dataset composed of SSDD, HRSID, and SAR-AIRcraft datasets, CCDN-DETR achieves a mean Average Precision (mAP) of 91.9%. Furthermore, it demonstrates significant competitiveness with 83.7% mAP on the multi-class MSAR dataset compared to CNN-based models.

show abstract

“…The research community has over the years built on these early efforts, moving from more traditional techniques [6]- [10] to deep learning based techniques. Several general trends can be observed: two-stage detectors [11]- [18], one-stage detectors [19]- [25], anchor free detectors based on different architectures such as CenterNet [26], [27], FCOS [28], [29] or other methods [30], [31]. Contextual information is leveraged to perform detection in complex scenes [32], [33].…”

Section: Introductionmentioning

confidence: 99%

Ship Detection From Raw SAR Echoes Using Convolutional Neural Networks

De Sousa,

Pilikos,

Azcueta

et al. 2024

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Synthetic Aperture Radar (SAR) is an indispensable tool for marine monitoring. Conventional data processing involves data down-linking and on-ground operations for image focusing, analysis and ship detection. These steps take significant amount of time, resulting in potentially critical delays. In this work, we propose a ship detection algorithm that operates directly on raw SAR echoes, based on convolutional neural networks. To evaluate our approach, we performed experiments using raw data simulations and real raw SAR data from Sentinel-1 stripmap mode scenes. Preliminary results on this set show the capability of detecting multiple ships from raw data with similar accuracy as using Single-Look-Complex (SLC) images as input. Simultaneously, running time is reduced significantly, by-passing the image focusing step. This illustrates the great potential of deep learning, moving towards more intelligent SAR systems.

show abstract

A Fast Threshold Neural Network for Ship Detection in Large-Scene SAR Images

Cited by 23 publications

References 34 publications

Lightweight CFARNets for Landmine Detection in Ultrawideband SAR

Lightweight CFARNets for Landmine Detection in Ultrawideband SAR

CCDN-DETR: A Detection Transformer Based on Constrained Contrast Denoising for Multi-Class Synthetic Aperture Radar Object Detection

Ship Detection From Raw SAR Echoes Using Convolutional Neural Networks

Contact Info

Product

Resources

About