Multiscale Feature Enhancement Network for Salient Object Detection in Optical Remote Sensing Images
Aircraft detection in synthetic aperture radar (SAR) images plays an essential role in satellite observation and military decisions. Due to discrete scattering properties, speckle noise interference, and various aircraft types, many existing methods struggle to achieve the desired detection performance. In this article, we propose an innovative semantic condition constraint guided feature aware network (SCFNet) for detecting different aircraft categories in SAR images. First, considering the discrete scattering properties of aircraft, we design a local-global feature aware module (LGA-M) and morphological-semantic feature aware module (MSF-M), which can effectively extract the fine-grained feature information contained in SAR images. Second, to effectively fuse different feature information, we construct a feature fusion pyramid (FFP), which uses different branches and paths to reasonably merge multiple feature information types and suppresses background information interference. Third, according to the structure characteristics of aircraft, the global coordinate attention mechanism (G-CAT) is presented to highlight foreground target features and suppress speckle noise interference. Finally, we construct semantic condition constraints, including constraint condition setting, semantic information calculation, and template matching, to improve aircraft localization and recognition accuracy. Extensive experiments demonstrate that the proposed SCFNet can obtain state-of-the-art performance on the SAR aircraft detection dataset, which achieves AP and F1 Score of 94.83% and 95.58%, respectively. The related implementation codes will be made publicly available at https://github.com/darkseid-arch/AirDetection.
Paper
Full text
Multiscale Feature Enhancement Network for Salient Object Detection in Optical Remote Sensing Images
OpenAlex · Visual Attention and Saliency Detection · 2022
Abstract
Salient object detection (SOD) in optical remote sensing images (RSIs) is a valuable and challenging task. Although many SOD methods for RSIs have been proposed, there are still some problems, such as insufficient feature extraction, relatively worse detection results for tiny objects, and serious interference by background clutter and lighting shadows. In this article, we propose a novel multiscale feature enhancement network (MFENet) for SOD in optical RSIs. Specifically, MFENet first uses the global feature perception (GFP) module to extract the multiscale and global feature information of salient object regions. Second, the global features are enhanced from the channel and spatial dimensions by using the mixed feature enhancement (MFE) module. Third, the semantic feature guidance (SFG) module is constructed to guide the model to obtain fine-grained semantic feature information. Finally, the boundary optimization module (BOM) is designed to refine the boundary features and obtain accurate saliency detection results. To evaluate the performance of MFENet, extensive experiments are conducted on the publicly available ORSSD, EORSSD, and ORSI-4199 datasets. The quantitative and qualitative results show that our method outperforms the existing state-of-the-art SOD methods in several evaluation metrics and exhibits better inference speed (FPS) and calculation parameters. The code and results of MFENet are available at https://github.com/darkseid-arch/MFENet.