检索结果-内蒙古大学图书馆

Lightweight vision Mamba for weather-degraded remote sensing image restoration

signal image AND VIDEO processing 2025年第1期19卷 1-9页

作者： Li, Yufeng Wu, Shuang Chen, Hongming Shenyang Aerosp Univ Coll Elect Informat Engn Shenyang 110136 Peoples R China Dalian Maritime Univ Coll Informat Sci & Technol Dalian 116026 Peoples R China

In the field of remote sensing, the acquired images are often severely degraded due to adverse weather conditions, such as haze and raindrops, posing significant challenges for subsequent visual tasks. Although CNN and Transformer have been widely applied to address these issues, they struggle to balance the relationship between global scene recovery, local detail preservation, and computational efficiency, leading to an imbalance between model performance and efficiency. To this end, we propose a lightweight and efficient visual state space model for remote sensing image restoration. Specifically, we propose the Efficient Vision Mamba Block as the core component of the model, incorporating the State Space Model to leverage its linear complexity for modeling long-range dependencies. Furthermore, we design a multi-router scanning strategy to perform global modeling of remote sensing images, capturing large spatial features from different routes and directions. Compared with existing methods that employ fixed-direction scanning, our approach avoids information redundancy caused by repeated scanning, making the model better adaptable to the complex and changeable weather conditions. Extensive experiments validate the superiority of our proposed model, outperforming state-of-the-art methods on both the StateHaze1k and UAV-Rain1k datasets.

关键词： remote sensing image restoration State space model Mamba

来源：评论

学校读者我要写书评

暂无评论

BAIDet: remote sensing image object detector based on background and angle information

引用

signal image AND VIDEO processing 2024年第12期18卷 9295-9304页

作者： Yu, Jiangfeng Sun, Lin Song, Shuhua Guo, Guolong Chen, Kai Shandong Univ Sci & Technol Coll Geodesy & Geomat Qingdao 266590 Peoples R China GEOVIS Wisdom Technol Co Ltd Qingdao 266100 Peoples R China Yangtze Univ Sch Geosci Wuhan 430000 Peoples R China Lingnan Normal Univ Sch Math & Stat Zhanjiang 524048 Peoples R China 91001 Troops Beijing 100309 Peoples R China

remote sensing object detection, with large differences in object size, arbitrary orientation and tight arrangement, leads to difficulties in object recognition and localization. Therefore, a remote sensing image object Detector (BAIDet) based on Background and Angle Information is proposed in this paper. Firstly, a large convolutional kernel global attention module is designed to fully utilize the global information of remote sensing images by expanding the receptive field. And obtain the edge information of ground objects through deformable convolution. Secondly, an angle-sensitive probabilistic intersection-over-union loss function (AS-ProbIoU Loss) is developed for bounding box regression for oriented object detection. Finally, experimental results on four remote sensing image datasets, DOTA, HRSC 2016, UCAS-AOD, and DIOR-R, demonstrated the effectiveness of this method.

关键词： Rotated object detection Deep convolution neural networks Angle loss Spatial attention

来源：评论

学校读者我要写书评

暂无评论

remote sensing image denoising based on deformable convolution and attention-guided filtering in progressive framework

引用

signal image AND VIDEO processing 2024年第11期18卷 8195-8205页

作者： Liu, Hualin Li, Zhe Lin, Shijie Cheng, Libo Changchun Univ Sci & Technol Sch Math & Stat Changchun 130022 Peoples R China Changchun Univ Sci & Technol Zhongshan Res Inst Lab Remote Sensing Technol & Big Data Anal Zhongshan 528437 Peoples R China

remote sensing image denoising tasks are challenged by complex noise distributions and multiple noise types, including a mixture of additive Gaussian white noise (AWGN) and impulse noise (IN). For better image recovery, complex contextual information needs to be balanced while maintaining spatial details. In this paper, a denoising model based on multilevel progressive image recovery is proposed to address the problem of remote sensing image denoising. In our model, the deformable convolution improves spatial feature sampling to effectively capture image details. Meanwhile, attention-guided filtering is used to pass the output images from the first and second stages to the third stage in order to prevent information loss and optimize the image recovery effect. The experimental results show that under the mixed noise scene of Gaussian and pepper noise, our proposed model shows superior performance relative to several existing methods in terms of both visual effect and objective evaluation indexes. Our model can effectively reduce the influence of image noise and recover more realistic image details.

关键词： Progressive framework Deformable convolution Attention-guided filtering Denoising

来源：评论

学校读者我要写书评

暂无评论

image and signal processing for remote sensing XXVIII

Image and Signal Processing for Remote Sensing XXVIII

引用

image and signal processing for remote sensing XXVIII 2022

ISBN: (纸本)9781510655379

The proceedings contain 33 papers. The topics discussed include: fast on-board ship detection in panchromatic images based on micro-nano satellite;star centroid extraction in the image sequences of a discrete detector array on geostationary orbit;two-step registration of near-space remote sensing images via deep neural networks;accuracy improvement with input image upscaling for in-orbit object detection;an automatic approach for ice-block falls detection in the Martian north polar region;compressed domain classification of remote sensing scene images based on sub-band data fusion;accelerating a two-stage object detector for high quality in-orbit remote sensing;fire image detection based on clustering data mining techniques;fusion of optical and SAR satellite data for environmental monitoring: assessment of damages and disturbances originated by forest fires;and multi-temporal change detection based on China’s domestic hyperspectral remote sensing satellite images.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Enhanced Swin Transformer and Edge Spatial Attention for remote sensing image Semantic Segmentation

引用

IEEE signal processing LETTERS 2025年 32卷 1296-1300页

作者： Liu, Fuxiang Hu, Zhiqiang Li, Lei Li, Hanlu Liu, Xinxin Beijing Inst Technol Minist Educ Key Lab Dynam & Control Flight Vehicle Beijing 100081 Peoples R China Chongqing Vocat Inst Tourism Chongqing 409099 Peoples R China North Nav Control Technol Co Ltd Beijing 100176 Peoples R China

Combining convolutional neural networks (CNNs) and transformers is a crucial direction in remote sensing image semantic segmentation. However, due to differences in the spatial information focus and feature extraction methods, existing feature transfer and fusion strategies do not effectively integrate the advantages of both approaches. To address these issues, we propose a CNN-transformer hybrid network for precise remote sensing image semantic segmentation. We propose a novel Swin Transformer block to optimize feature extraction and enable the model to handle remote sensing images of arbitrary sizes. Additionally, we design an Edge Spatial Attention module to focus attention on local edge structures, effectively integrating global features and local details. This facilitates efficient information flow between the Transformer encoder and CNN decoder. Finally, a multi-scale convolutional decoder is employed to fully leverage both global information from the Transformer and local features from the CNN, leading to accurate segmentation results. Our network achieved state-of-the-art performance on the Vaihingen and Potsdam datasets, reaching mIoU and F1 scores of 67.37% and 79.82%, as well as 72.39% and 83.68%, respectively.

关键词： Transformers image edge detection Feature extraction Convolution Vegetation Semantic segmentation Decoding remote sensing Data mining Training Edge detection remote sensing image semantic segmentation Swin transformer

来源：评论

学校读者我要写书评

暂无评论

Multi-stage remote sensing super-resolution network with deep fusion and structure enhancement based on CNN and transformer

引用

signal image AND VIDEO processing 2025年第5期19卷 1-10页

作者： Liu, Jingyi Yang, Xiaomin Sichuan Univ Coll Elect & Informat Engn Chengdu 610065 Sichuan Peoples R China Hubei Minzu Univ Coll Intelligent Syst Sci & Engn Enshi 445000 Hubei Peoples R China Hubei Minzu Univ Key Lab Green Mfg Superlight Elastomer Mat State Ethn Affairs Commiss Enshi 445000 Hubei Peoples R China

remote sensing super-resolution remains a research hotspot in the field of remote sensing. Unlike natural images, remote sensing images typically possess rich scenes and complex spatial structures, making it challenging to enhance their resolution. This paper introduces a hybrid attention module that integrates CNNs, Transformers, and spatial attention to effectively extract multi-level features from remote sensing images. The module seamlessly integrates the early feature projection processes of CNNs and Transformers, harmoniously blending CNNs' strength in local modeling with Transformers' expertise in global modeling, thereby simplifying the complexity of fusion. To accurately preserve the intricate details of the input images, we construct a structure enhancement module that focuses on extracting edge and linear structural details to retain the complex features of the images. Furthermore, we build a multi-level remote sensing super-resolution network. A series of experiments conducted on the AID dataset demonstrate its excellent effectiveness and generalization capabilities.

关键词： remote sensing Super-resolution CNN Transformer Structure enhancement

来源：评论

学校读者我要写书评

暂无评论

CFFormer: Channel Fourier Transformer for remote sensing Super Resolution

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND remote sensing 2025年 18卷 569-583页

作者： Xie, Zexin Wang, Jian Song, Wei Du, Yanling Xu, Huifang Yang, Qinhan Shanghai Ocean Univ Coll Informat Technol Shanghai 201308 Peoples R China Shanghai Jian Qiao Univ Coll Informat Technol Shanghai 201315 Peoples R China

The objective of super-resolution in remote sensing imagery is to enhance low-resolution images to recover high-quality details. With the rapid progress of deep learning technology, the deep learning-based super-resolution technology for remote sensing images has also made remarkable achievements. However, these methods encounter several challenges. They often struggle with processing long-range spatial information that encompasses complex scene changes, adversely affecting the image's coherence and accuracy. Furthermore, the lack of connectivity in feature extraction blocks hinders effective feature utilization in deeper network layers, leading to issues such as gradient vanishing and exploding. Additionally, constraints in the spatial domain of previous methods frequently result in severe shape distortion and blurring. To address these issues, this study proposes the CFFormer, a new super-resolution framework that employs the Swin Transformer as its core architecture and incorporates the Channel Fourier Block (CFB) to refine features in the frequency domain. The Global Attention Block (GAB) is also integrated to enhance global information capture, thereby improving the extraction of spatial features. To increase model stability and feature utilization efficiency, a Jump-Joint Fusion Mechanism is designed, culminating in a Residual Fusion Swin Transformer Block (RFSTB) that alleviates the gradient vanishing issue and optimizes feature reuse. Experimental results confirm the CFFormer's superior performance in remote sensing image reconstruction, demonstrating outstanding perceptual quality and reliability. Notably, the CFFormer achieves a Peak signal-to-Noise Ratio (PSNR) of 29.83 dB on the UcMercedx4 dataset, surpassing the SwinIR method by approximately 0.5 dB, indicating a substantial enhancement.

关键词： Superresolution remote sensing Spatial resolution Feature extraction Fast Fourier transforms Deep learning Current transformers image reconstruction Convolutional neural networks Frequency-domain analysis Convolutional neural network (CNN) fast Fourier transform (FFT) remote sensing image processing super resolution (SR) transformer

来源：评论

学校读者我要写书评

暂无评论

UNSUPERVISED remote sensing HAZE REMOVAL BASED ON SALIENCY-GUIDED TRANSMISSION REFINEMENT 49

UNSUPERVISED REMOTE SENSING HAZE REMOVAL BASED ON SALIENCY-G...

引用

49th IEEE International conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Zheng, Ruohui Zhang, Libao Beijing Normal Univ Sch Artificial Intelligence Beijing 100875 Peoples R China

ISBN: (纸本)9798350344868;9798350344851

Haze causes information loss and quality degradation in remote sensing images. Unsupervised learning-based dehazing methods aim to reduce reliance on paired hazy images and their labels. However, complex mapping relationships often increase the difficulty in network convergence, resulting in color distortion and loss of texture details in remote sensing images. To address these issues, we propose an unsupervised haze removal method based on saliency-guided transmission refinement for remote sensing images. Firstly, we introduce a saliency-guided transmission refinement method, which decomposes and recombines two transmission maps obtained under different conditions, guided by saliency information. Secondly, we propose a loss function comprising energy loss and texture loss. The energy loss provides an energy reference based on the coarse transmission estimation, while the texture loss enhances the preservation of texture details. Experimental results demonstrate that our method achieves comparable performance to several supervised methods.

关键词： image restoration image enhancement remote sensing dehazing unsupervised learning

来源：评论

学校读者我要写书评

暂无评论

Effective Denoising of InSAR Phase images via Compressive sensing

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND remote sensing 2024年 17卷 17772-17779页

作者： Kang, Min-Seok Baek, Jae-Min Kongju Natl Univ Div Elect Elect & Control Engn Cheonan 31080 South Korea Gangneung Wonju Natl Univ Dept Mech Engn Wonju 26403 South Korea

Interferometric synthetic aperture radar (InSAR) denoising is an essential processing step in deformation measurement and topography reconstruction. A noisy InSAR phase image gives rise to the phase unwrapping difficulties and even results in the degradation of various final products of InSAR. To address this issue, we develop a compressive sensing (CS)-based InSAR phase denoising technique in this article. Since the spectrum of the InSAR phase image is usually sparse in the 2-D frequency domain, the estimation of sensing dictionary matrix of the linear system between the InSAR phase signal and its spectrum in the pursuit of sparsity is considered for InSAR phase denoising. The optimization problem derived by the signal parameterization approach is effectively carried out by estimating the basis function that is closely analogous to the strongest signal component in the spectrum of the InSAR phase image. The proposed method is effectively capable of eliminating noise and preserving detailed fringe information of InSAR. In the end, simulations and experimental results demonstrate that the proposed scheme outperforms other conventional InSAR phase denoising methods.

关键词： Vectors Phase measurement Noise reduction image reconstruction Sensors Radar imaging Phase noise Compressive sensing (CS) interferometric synthetic aperture radar (InSAR) phase filtering sparse signal processing synthetic aperture radar (SAR)

来源：评论

学校读者我要写书评

暂无评论

SEMANTIC SEGMENTATION FOR MULTI-SCENE remote sensing imageS WITH NOISY LABELS BASED ON UNCERTAINTY PERCEPTION 49

SEMANTIC SEGMENTATION FOR MULTI-SCENE REMOTE SENSING IMAGES ...

引用

49th IEEE International conference on Acoustics, Speech, and signal processing (ICASSP)

作者： Lyu, Xinran Zhang, Libao Beijing Normal Univ Sch Artificial Intelligence Beijing 100875 Peoples R China

ISBN: (纸本)9798350344868;9798350344851

As the annotation of remote sensing images requires domain expertise, it is difficult to construct a large-scale and accurate annotated dataset. image-level annotation data learning has become a research hotspot. In addition, due to the difficulty in avoiding mislabeling, label noise cleaning is also a concern. In this paper, a semantic segmentation method for remote sensing images based on uncertainty perception with noisy labels is proposed. The main contributions are three-fold. First, a label cleaning method based on iterative learning is presented to handle noise labels such as missing or incorrect annotations. Second, a two-stage semantic segmentation model is proposed for image-level annotation, which eliminates the need for post-processing steps during testing. Lastly, a complementary uncertainty perception function is introduced to improve the utilization of dataset features and enhance the accuracy of segmentation. The effectiveness of this method was verified through comprehensive evaluation with 7 models on four datasets.

关键词： remote sensing weak annotation label noise cleaning iterative learning uncertainty perception

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：