Article
Geochemistry & Geophysics
Xiangtao Zheng, Wenjing Chen, Xiaoqiang Lu
Summary: This study introduces a novel network architecture that can simultaneously explore the spatial and spectral information of multispectral images, leading to the reconstruction of hyperspectral images.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Geochemistry & Geophysics
Pengfei Liu
Summary: This letter proposes an adaptive pansharpening model for the fusion of multispectral and panchromatic images, incorporating spatial and spectral anisotropic tensor total variation for spatial gradient and spectral feature preservation.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Geochemistry & Geophysics
Bowen Zhao, Jiesi Zheng, Yafei Dong, Ning Shen, Jiangxin Yang, Yanlong Cao, Yanpeng Cao
Summary: This article proposes a pseudo-panchromatic image (PPI) edge-infused spatial-spectral adaptive residual network (PPIE-SSARN) for MSFA image demosaicing. The method compensates for the spatial and spectral differences of reconstructed multispectral images and enriches the edge-related information using a two-branch model. Experimental results demonstrate the superiority of the proposed method in spatial accuracy and spectral fidelity. The models and code will be publicly available.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Optics
Ningning LI, Qiuhua Tang, Yilan Chen, Zhipeng Dong, J. I. E. LI, Xuancheng Fu
Summary: Satellite derived bathymetry (SDB) is a significant and cost-effective method for obtaining shallow seabed topography. This study proposes an SDB approach incorporating spectral and spatial information of multispectral images to improve the accuracy of bathymetry inversion. Experimental results show that the method effectively reduces the error in bathymetry estimation caused by spatial heterogeneity of the seabed.
Article
Environmental Sciences
Qingwang Wang, Zifeng Zhang, Xueqian Chen, Zhifeng Wang, Jian Song, Tao Shen
Summary: This paper proposes a deep spatial graph convolution network with adaptive spectral aggregated residuals (DSGCN-ASR) for classifying multispectral point clouds. The method effectively overcomes the limitations of existing methods in information aggregation and spatial-spectral information utilization, resulting in improved model performance.
Article
Remote Sensing
Nan Chen, Lichun Sui, Biao Zhang, Hongjie He, Kyle Gao, Yandong Li, Jose Marcato Junior, Jonathan Li
Summary: A novel fusion method for hyperspectral images and multispectral images is proposed, which overcomes limitations of existing methods and achieves state-of-the-art performance in urban green infrastructure monitoring.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2021)
Article
Geochemistry & Geophysics
Tianshuai Li, Tianzhu Liu, Yukun Wang, Xian Li, Yanfeng Gu
Summary: In this article, a multitemporal spectral reconstruction network (MTSRN) is proposed to reconstruct hyperspectral (HS) images from multitemporal multispectral (MS) images. By extracting temporal features and utilizing a multitemporal fusion network, better HS data can be obtained.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Computer Science, Theory & Methods
Yong Chen, Taoshun He
Summary: This paper introduces an adaptive weighted anisotropic diffusion model for image denoising, combining a patch-based diffusivity function with a local diffusivity function. A variable time step is designed to address the problem of over-smoothness. Experimental results demonstrate that the proposed model outperforms some representative anisotropic diffusion models in terms of both quantitative metrics and visual performance.
MULTIDIMENSIONAL SYSTEMS AND SIGNAL PROCESSING
(2021)
Article
Geochemistry & Geophysics
Jingyi Cai, Wei He, Hongyan Zhang
Summary: This article proposed an HSI denoising and destriping method based on anisotropic spatial and spectral total variation regularized double LR approximation (ATVDLR), which can achieve superior performance in complex mixed noise and image structural information protection.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Remote Sensing
Shuo Shi, Biwu Chen, Sifu Bi, Junkai Li, Wei Gong, Jia Sun, Bowen Chen, Lin Du, Jian Yang, Qian Xu, Fei Wang, Shalei Song
Summary: Multispectral LiDAR introduces a new data type, multispectral point cloud, which improves classification performance by separating spatial and spectral information. This study proposes a spatial-spectral classification framework with four steps: neighborhood selection, feature extraction and selection, classification, and label smoothing. Experimental results demonstrate the effectiveness of this method in selecting precise neighborhoods, extracting effective features, and refining classification results.
GEO-SPATIAL INFORMATION SCIENCE
(2023)
Article
Computer Science, Information Systems
Rouzbeh Shad, Seyyed Tohid Seyyed-Al-hosseini, Yaser Maghsoodi Mehrani, Marjan Ghaemi
Summary: Previous studies have not yet found a single attribute with the highest accuracy for different applications. This paper proposes a novel classification system using Support Vector Machine (SVM) that has the highest strength against possible noises. The performance of this system is evaluated on selected satellite images, and the results show that the proposed method outperforms other techniques in terms of accuracy.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Geochemistry & Geophysics
Weiwei Sun, Kai Ren, Xiangchao Meng, Gang Yang, Jiangtao Peng, Jiancheng Li
Summary: This study proposes a method for reconstructing remote sensing images with high temporal, spatial, and spectral resolution by fusing the temporal, spatial, and spectral information from multiple sources of remote sensing images. By utilizing tensor subspace decomposition and reconstruction networks, it effectively utilizes low-resolution hyperspectral images and high-resolution multispectral images. Experimental results demonstrate that the method achieves high-quality fusion results, exhibits comparable performance, and has robustness and practicality.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2023)
Article
Environmental Sciences
Erdem Ozer, Ugur Murat Leloglu
Summary: Wetlands are important for biodiversity and ecosystems, and monitoring their extent and ground characteristics is crucial for sustainable development. This study proposes a remote sensing-based method to determine the extent of wetlands and extract their ground features at the sub-pixel level. Experimental results using wetlands in Turkey demonstrate the accuracy of the method in determining wetland extent and identifying land cover classes.
GEOCARTO INTERNATIONAL
(2022)
Article
Computer Science, Theory & Methods
Yong Chen
Summary: This paper proposes a robust anisotropic diffusion filter that can simultaneously remove additive white Gaussian noise and impulsive noise, using a robust spatial gradient estimator to separate features and noise. Experimental results demonstrate that the proposed filter outperforms benchmark models in both quantitative metrics and visual performance.
MULTIDIMENSIONAL SYSTEMS AND SIGNAL PROCESSING
(2022)
Article
Geography, Physical
Chunyu Zhu, Shangqi Deng, Jiaxin Li, Ying Zhang, Liwei Gong, Liangbo Gao, Na Ta, Shengbo Chen, Qiong Wu
Summary: This study introduces SwinGAN, a fusion network combining Swin Transformer, CNN, and GAN architectures, to improve data resolution in hyperspectral remote sensing image (HSI) fusion with multispectral remote sensing images (MSI). SwinGAN employs a detail injection framework to separately extract HSI and MSI features, fusing them to generate spatial residuals. These residuals are injected into the supersampled HSI to produce the final image, while a pure CNN architecture acts as the discriminator, enhancing the fusion quality. Additionally, a new adaptive loss function is introduced to improve image fusion accuracy, using L1 loss as the content loss and introducing spatial and spectral gradient loss functions.
INTERNATIONAL JOURNAL OF DIGITAL EARTH
(2023)
Article
Computer Science, Artificial Intelligence
Xiaoxiong Zheng, Tao Chen
Summary: Semantic segmentation technology is crucial for interpreting remote sensing images. Traditional methods cannot accurately segment high spatial resolution images. This paper explores the use of U-Net algorithm in remote sensing image classification and segmentation, achieving high accuracy and credibility through a neighborhood voting method for uncertain pixels.
NEURAL COMPUTING & APPLICATIONS
(2023)
Article
Environmental Sciences
Zhice Fang, Yi Wang, Gonghao Duan, Ling Peng
Summary: This study introduces a new ensemble framework to predict landslide susceptibility by combining decision trees with the rotation forest technique. By selecting training and validation sets based on historical landslide locations, screening landslide conditioning factors, producing training subsets, and integrating all DTs classification results using RF ensemble technique, the framework effectively improves the spatial prediction of landslides. The experimental results show that the proposed ensemble methods outperform traditional DTs and other popular ensemble methods in terms of predictive values.
Article
Environmental Sciences
Lingran Zhao, Ruiqing Niu, Bingquan Li, Tao Chen, Yueyue Wang
Summary: The traditional manual interpretation method for mine remote sensing pre-survey is subjective and time-consuming. To improve the efficiency and reduce labor costs, two improved instance segmentation models are proposed. The evaluation on different satellite image datasets shows that the improved models outperform the traditional method in mine detection tasks.
Article
Computer Science, Theory & Methods
Chen Wang, Tao Chen, Antonio Plaza
Summary: This study introduces a new approach using a mining feature-enhanced ResNet network to obtain effective image features and enhance the extraction of open-pit and waste-dump areas in mining areas. The results show that the network achieves the best overall accuracy in both large and small study areas.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Environmental Sciences
Renxiang Huang, Tao Chen
Summary: Efficient and accurate landslide recognition is crucial for disaster prevention and post-disaster rescue efforts. We proposed a novel knowledge distillation network based on Swin-Transformer (Distilled Swin-Transformer, DST) to tackle the challenges of long model runtimes and inefficiency in deep learning approaches. Our model achieved improved performance in landslide recognition by combining remote sensing images (RSIs) with nine landslide influencing factors (LIFs). Compared to other neural networks, DST showed the best overall accuracy (98.1717%) and required the lowest number of floating point operations (FLOPs).
Article
Environmental Sciences
Na Lin, Hailin Quan, Jing He, Shuangtao Li, Maochi Xiao, Bin Wang, Tao Chen, Xiaoai Dai, Jianping Pan, Nanjie Li
Summary: Urban vegetation plays a crucial role in the urban ecological system, and efficient extraction of vegetation information is important. Therefore, this study proposes a new model called SD-UNet, which improves the accuracy and generalization ability through the introduction of dense connections and separable convolutions.
Article
Environmental Sciences
Na Lin, Di Zhang, Shanshan Feng, Kai Ding, Libing Tan, Bin Wang, Tao Chen, Weile Li, Xiaoai Dai, Jianping Pan, Feifei Tang
Summary: This study integrates Shapley Additive Explanation (SHAP) and Optuna (OPT) hyperparameter tuning into four basic machine learning algorithms and applies them to landslide extraction in Fengjie County, Chongqing, China. The experimental results show that the four SHAP-OPT models have an accuracy above 92% and a training time less than 1.3 seconds. Among them, SHAP-OPT-XGBoost achieves the highest accuracy (96.26%) and can extract landslide distribution information accurately and quickly.
Article
Geochemistry & Geophysics
Tao Chen, Yang Liu, Yuxiang Zhang, Bo Du, Antonio Plaza
Summary: Sparse unmixing (SU) is a method widely used for interpreting remotely sensed hyperspectral images. It avoids the need to extract pure signatures and directly selects spectra from a known library. However, SU generally lacks spatial information, which can be addressed by utilizing low-rank and sparse features in local regions.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)
Article
Engineering, Electrical & Electronic
Tao Chen, Xiaoxiong Zheng, Ruiqing Niu, Antonio Plaza
Summary: This article proposes a hybrid open-pit mining mapping framework that utilizes high-resolution satellite images and an improved neural network to improve the accuracy of mapping open-pit mining areas. By comparing with other methods, the framework demonstrates better performance in various accuracy metrics.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2022)
Article
Engineering, Electrical & Electronic
Tao Chen, Zhiyuan Lu, Yue Yang, Yuxiang Zhang, Bo Du, Antonio Plaza
Summary: This article introduces a new algorithm for remote sensing image change detection, which combines attention mechanism with UNet and has performed well in solving complex change detection problems, especially in weak change detection and noise suppression.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2022)
Article
Engineering, Electrical & Electronic
Lulu Wan, Tao Chen, Antonio Plaza, Haojie Cai
Summary: Hyperspectral unmixing involves obtaining endmembers and abundance vectors through linear or nonlinear models. A one-dimensional convolutional neural network (CNN) is proposed for supervised unmixing, showing effectiveness and stability in comparison to traditional linear unmixing algorithms. The CNN-based method outperforms other methods in terms of RMSE on both simulated and real datasets.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Engineering, Electrical & Electronic
Tong Liu, Tao Chen, Ruiqing Niu, Antonio Plaza
Summary: In this study, an accurate Landslide Detection Mapping (LDM) model was constructed based on convolutional neural networks, residual neural networks, and DenseNets, considering ZY-3 high spatial resolution data and conditioning factors. The experimental results demonstrated that these models performed well, with accuracy above 0.95. DenseNet incorporating RS images and CFs outperformed other models in terms of evaluation metrics, with improvements in Kappa coefficients and ACC. Elevation factor was identified as the most important factor in the landslide model construction experiment.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Engineering, Electrical & Electronic
Yuxiang Zhang, Yanni Dong, Ke Wu, Tao Chen
Summary: The proposed method in this article, based on an Otsu-based isolation forest, effectively separates anomalies from backgrounds by assembling multiple binary trees and using the Otsu-based splitting criterion for a more discriminative binary tree construction.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Engineering, Electrical & Electronic
Haojie Cai, Tao Chen, Ruiqing Niu, Antonio Plaza
Summary: This study addresses the issues of insufficient accuracy, scarcity of samples, and low efficiency in landslide mapping by utilizing DenseNets and a new sample library. By incorporating environmental factors and optimizing DenseNet, the accuracy of detecting landslides improves significantly, showing promising applicability in large-scale landslide identification scenarios.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Geochemistry & Geophysics
Guoli Li, Shuang Liu, Xiange Jian, Dan Zhu, Lihua Fu, Tao Chen, Xiangyun Hu
Summary: Identification of lineament structure is crucial for determining metallogenic areas and geologic structure distribution. The use of edge detection methods with gravity, magnetic, and remote sensing data can provide more geological information. New edge detectors for potential field derivatives have advantages in producing clear edges, equalizing anomalies, and distinguishing superimposed anomalies. Combining different data types to extract lineaments results in a more comprehensive interpretation of lineament structures.
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING
(2022)