SAR Image Aircraft Target Recognition Based on Improved YOLOv5

Wang, Xing; Hong, Wen; Liu, Yunqing; Hu, Dongmei; Xin, Ping

doi:10.3390/app13106160

Cited by 5 publications

(1 citation statement)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This layer has a broader receptive field, which covers a larger feature area, thereby enhancing the detection performance for large-sized targets. The FPN + PAN structure, through the application of a progressive fusion strategy, effectively integrates multi-scale information, thereby enhancing the model's detection capabilities across various object sizes [37]. The small object feature layer is designed to capture subtle details that are appropriate for detecting small objects.…”

Section: Feature Pyramid Neck Network: Fpn + Panmentioning

confidence: 99%

MSSD-Net: Multi-Scale SAR Ship Detection Network

Wang,

Xu,

Huang

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

In recent years, the development of neural networks has significantly advanced their application in Synthetic Aperture Radar (SAR) ship target detection for maritime traffic control and ship management. However, traditional neural network architectures are often complex and resource intensive, making them unsuitable for deployment on artificial satellites. To address this issue, this paper proposes a lightweight neural network: the Multi-Scale SAR Ship Detection Network (MSSD-Net). Initially, the MobileOne network module is employed to construct the backbone network for feature extraction from SAR images. Subsequently, a Multi-Scale Coordinate Attention (MSCA) module is designed to enhance the network’s capability to process contextual information. This is followed by the integration of features across different scales using an FPN + PAN structure. Lastly, an Anchor-Free approach is utilized for the rapid detection of ship targets. To evaluate the performance of MSSD-Net, we conducted extensive experiments on the Synthetic Aperture Radar Ship Detection Dataset (SSDD) and SAR-Ship-Dataset. Our experimental results demonstrate that MSSD-Net achieves a mean average precision (mAP) of 98.02% on the SSDD while maintaining a compact model size of only 1.635 million parameters. This indicates that MSSD-Net effectively reduces model complexity without compromising its ability to achieve high accuracy in object detection tasks.

show abstract

Section: Feature Pyramid Neck Network: Fpn + Panmentioning

confidence: 99%

MSSD-Net: Multi-Scale SAR Ship Detection Network

Wang,

Xu,

Huang

et al. 2024

Remote Sensing

View full text Add to dashboard Cite

show abstract

Aircraft detection in high-resolution synthetic aperture radar images based on optical training dataset

Cui,

Peng,

Dong

et al. 2024

Infrared, Millimeter-Wave, and Terahertz Technologies XI

View full text Add to dashboard Cite

Synthetic Aperture Radar (SAR) emits microwave electromagnetic pulses and detects remote targets through the backscattered echo signals. With the continuous advancement of technology, high-resolution SAR imaging can accurately observe various targets with small sizes, dense connections, and diverse shapes. The steep terrain or tall objects in SAR images display prominent shadows due the obstruction of ground-facing electromagnetic waves resulting in weak echo signals in the shadowed areas. Small targets can generate discrete shadows with contours similar to optical contours in high-resolution SAR images when the radar observing angle is appropriate. This characteristic, which shares similarities with the target contour, can be utilized to build the correlation between SAR images and other modal images such as optical images. This study uses aircraft as an example to validate the feasibility of this approach. It trains the YOLOv5 object detection network on the Remote Sensing Object Detection (RSOD) dataset of optical airport images, which is then utilized to detect the shadows of aircraft in high-resolution SAR images, indirectly achieving aircraft detection.Firstly, the shadows in SAR images have an inverse relationship with the signal of the aircraft body in optical images in terms of grayscale. Therefore, it is possible to simply invert the grayscale of one of them. In this study, SAR images were chosen for grayscale inversion. After a simple grayscale inversion, the network detected a significant number of aircraft in the image at a confidence level of 0.2, while only part of aircraft were detected in the image without inversion. Besides, a series of adjustments were made to the brightness and contrast of the grayscale inverted image in order to find the optimal setting for aircraft detection. After scanning and adjusting the brightness and contrast, the network detects a certain number of additional aircraft in the grayscale inverted images at a confidence level of 0.2.The maximum number of aircraft detections was achieved at a specific filtering spatial frequency after applying filtering with different spatial frequencies. The overall detection result achieved an accuracy of over 80%.The maximum number of aircraft detections was achieved at a specific filtering spatial frequency after applying filtering with different spatial frequencies. The overall detection result achieved an accuracy of over 80%.

show abstract

Aircraft Target Interpretation Based on SAR Images

Wang,

Hong,

Liu

et al. 2023

Applied Sciences

Self Cite

View full text Add to dashboard Cite

The focus of the current research is how to effectively identify aircraft targets in SAR images. There are only 2000 SAR images and 6556 aircraft instances in our dataset. SAR images have complex backgrounds, and the sizes of aircraft targets are multi-scale. How to improve the detection accuracy of aircraft targets is the research topic of this paper, especially for small target detection. We proposed four improved methods based on YOLOv5s. Firstly, this paper proposed the structure of the multi-scale receptive field and channel attention fusion. It is applied at the shallow layer of the backbone of YOLOv5s. It can adjust the weights of the multi-scale receptive field during the training process to enhance the extraction ability of feature information. Secondly, we proposed four decoupled detection heads to replace the original part in YOLOv5s. It can improve the efficiency and accuracy of SAR image interpretation for small targets. Thirdly, in the case of the limited amount of SAR images, this paper proposed multiple data-augmentation methods, which can enhance the diversity and generalization of the network. Finally, this paper proposed the K-means++ to replace the original K-means to improve the network convergence speed and detection accuracy. Experiments demonstrate that the improved YOLOv5s can enhance the accuracy of SAR image interpretation by 9.3%, and the accuracy of small targets is improved more obviously, reaching 13.1%.

show abstract

SAR Image Aircraft Target Recognition Based on Improved YOLOv5

Cited by 5 publications

References 44 publications

MSSD-Net: Multi-Scale SAR Ship Detection Network

MSSD-Net: Multi-Scale SAR Ship Detection Network

Aircraft detection in high-resolution synthetic aperture radar images based on optical training dataset

Aircraft Target Interpretation Based on SAR Images

Contact Info

Product

Resources

About