Egyptian Journal of Remote Sensing and Space Sciences最新文献

Integration between spatial analysis with MCDA for best site selection of hospitals: A case study of Port Said Governorate, Egypt 将空间分析与MCDA相结合用于医院最佳选址：以埃及塞得港省为例

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-12-11 DOI: 10.1016/j.ejrs.2025.11.005

Hisham Nabil, Mahmoud El-Mewafi, Mohamed Zhran

The recognition and prioritization of hospital locations plays an important role in shaping the healthcare framework of any nation. The process of selecting suitable sites is an intricate, multi-faceted decision-making endeavor involving various stakeholders with their unique interests. The main objective of this study is to ascertain the most appropriate approach for the selection of the best location for the new hospital. Multi-criteria decision analysis (MCDA) presents a promising solution due to the multitude of criteria integral to the decision-making process. Accordingly, the current research focuses on four criteria (population density, proximity to road, distance to existing hospitals, and slope). This study employs three MCDA techniques, namely, analytical hierarchy process (AHP), fuzzy AHP (FAHP), and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). Additionally, MCDA techniques utilizing geographical information systems (GIS), applied to Port Said Governorate, Egypt, are introduced. The results obtained through the application of AHP, FAHP, and TOPSIS are carefully compared, encompassing criterion rankings and suggested hospital site locations. These results indicate that AHP, FAHP, and TOPSIS give the same result for the best site selection for the new hospital in the study area. The results of this study provide decision-makers and urban planners vital information that makes it easier to pinpoint the best site location for a new hospital.

医院位置的识别和优先排序在塑造任何国家的医疗保健框架方面都起着重要作用。选择合适地点的过程是一个复杂的、多方面的决策过程，涉及不同利益相关者的独特利益。本研究的主要目的是确定为新医院选择最佳位置的最合适方法。多准则决策分析（MCDA）提出了一个有前途的解决方案，由于众多的标准组成的决策过程。因此，目前的研究重点是四个标准（人口密度、靠近道路、与现有医院的距离和坡度）。本研究采用三种MCDA技术，即分析层次分析法（AHP）、模糊层次分析法（FAHP）和理想解相似偏好排序法（TOPSIS）。此外，还介绍了在埃及塞得港省应用的利用地理信息系统（GIS）的MCDA技术。通过应用AHP、FAHP和TOPSIS获得的结果进行了仔细的比较，包括标准排名和建议的医院选址。这些结果表明，AHP、FAHP和TOPSIS对研究区新医院的最佳选址给出了相同的结果。这项研究的结果为决策者和城市规划者提供了重要的信息，使他们更容易确定新医院的最佳选址。

{"title":"Integration between spatial analysis with MCDA for best site selection of hospitals: A case study of Port Said Governorate, Egypt","authors":"Hisham Nabil, Mahmoud El-Mewafi, Mohamed Zhran","doi":"10.1016/j.ejrs.2025.11.005","DOIUrl":"10.1016/j.ejrs.2025.11.005","url":null,"abstract":"<div><div>The recognition and prioritization of hospital locations plays an important role in shaping the healthcare framework of any nation. The process of selecting suitable sites is an intricate, multi-faceted decision-making endeavor involving various stakeholders with their unique interests. The main objective of this study is to ascertain the most appropriate approach for the selection of the best location for the new hospital. Multi-criteria decision analysis (MCDA) presents a promising solution due to the multitude of criteria integral to the decision-making process. Accordingly, the current research focuses on four criteria (population density, proximity to road, distance to existing hospitals, and slope). This study employs three MCDA techniques, namely, analytical hierarchy process (AHP), fuzzy AHP (FAHP), and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). Additionally, MCDA techniques utilizing geographical information systems (GIS), applied to Port Said Governorate, Egypt, are introduced. The results obtained through the application of AHP, FAHP, and TOPSIS are carefully compared, encompassing criterion rankings and suggested hospital site locations. These results indicate that AHP, FAHP, and TOPSIS give the same result for the best site selection for the new hospital in the study area. The results of this study provide decision-makers and urban planners vital information that makes it easier to pinpoint the best site location for a new hospital.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"29 1","pages":"Pages 1-14"},"PeriodicalIF":4.1,"publicationDate":"2025-12-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145718860","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Comparative assessment of GIS-based multi-criteria decision analysis (AHP) and machine learning (MaxEnt) approaches for wildfire susceptibility modeling in Montenegro 基于gis的多准则决策分析（AHP）和机器学习（MaxEnt）方法在黑山野火易感性建模中的比较评估

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-12-01 DOI: 10.1016/j.ejrs.2025.11.003

Filip Vujović , Goran Grozdanić , Radovan Đurović , Aleksandar Valjarević , Ivica Milevski

This study provides a national-scale assessment of wildfire susceptibility in Montenegro by comparing two modeling frameworks: a Geographic Information System–based Multi-Criteria Decision Analysis (GIS-MCDA) using fuzzy standardization with the Analytic Hierarchy Process (AHP), and a machine-learning approach based on the Maximum Entropy model (MaxEnt). Historical wildfire occurrences from the Moderate Resolution Imaging Spectroradiometer (MODIS) archive (2001–2024) were analyzed together with initial thirteen geoenvironmental initial causal criteria spanning vegetation, climatic, topographic and anthropogenic factors. After the Variance Inflation Factor (VIF) analysis, eleven causal criteria were retained for further modeling. Results of validation using receiver operating characteristic (ROC) curves showed clear performance differences: GIS-MCDA with fuzzy standardization and AHP weighting achieved low predictive accuracy (area under the curve, AUC = 0.51), whereas MaxEnt performed strongly (AUC = 0.81). These findings highlight the ability of MaxEnt to capture nonlinear relationships and complex interactions among geoenvironmental causal criteria, whereas GIS-MCDA with fuzzy standardization and AHP proved inadequate for reliable wildfire susceptibility assessment in this context. These results confirm the findings of previous studies, while showing an even weaker performance of GIS-MCDA with fuzzy standardization and AHP compared to earlier studies. Importantly, this is the first application of machine learning through MaxEnt for wildfire susceptibility assessment in Montenegro, providing a spatial basis for wildfire management and a foundation for national-scale wildfire risk assessment.

本研究通过比较两种建模框架：基于地理信息系统的多标准决策分析（GIS-MCDA）和基于最大熵模型（MaxEnt）的机器学习方法，对黑山的野火易感性进行了全国范围的评估。利用2001-2024年MODIS档案资料，结合植被、气候、地形和人为因素等13个地质环境初始因果准则，对历史野火发生情况进行了分析。在方差膨胀因子（VIF）分析后，保留了11个因果标准以进一步建模。使用受试者工作特征（ROC）曲线进行验证的结果显示出明显的性能差异：采用模糊标准化和AHP加权的GIS-MCDA的预测精度较低（曲线下面积，AUC = 0.51），而MaxEnt的预测精度较高（AUC = 0.81）。这些发现强调了MaxEnt能够捕捉地质环境因果标准之间的非线性关系和复杂相互作用，而具有模糊标准化和层次分析法的GIS-MCDA在这种情况下不足以进行可靠的野火易感性评估。这些结果证实了之前的研究结果，但与之前的研究相比，模糊标准化和层次分析法的GIS-MCDA的表现更弱。重要的是，这是通过MaxEnt首次将机器学习应用于黑山的野火易感性评估，为野火管理提供了空间基础，为全国范围的野火风险评估奠定了基础。

{"title":"Comparative assessment of GIS-based multi-criteria decision analysis (AHP) and machine learning (MaxEnt) approaches for wildfire susceptibility modeling in Montenegro","authors":"Filip Vujović , Goran Grozdanić , Radovan Đurović , Aleksandar Valjarević , Ivica Milevski","doi":"10.1016/j.ejrs.2025.11.003","DOIUrl":"10.1016/j.ejrs.2025.11.003","url":null,"abstract":"<div><div>This study provides a national-scale assessment of wildfire susceptibility in Montenegro by comparing two modeling frameworks: a Geographic Information System–based Multi-Criteria Decision Analysis (GIS-MCDA) using fuzzy standardization with the Analytic Hierarchy Process (AHP), and a machine-learning approach based on the Maximum Entropy model (MaxEnt). Historical wildfire occurrences from the Moderate Resolution Imaging Spectroradiometer (MODIS) archive (2001–2024) were analyzed together with initial thirteen geoenvironmental initial causal criteria spanning vegetation, climatic, topographic and anthropogenic factors. After the Variance Inflation Factor (VIF) analysis, eleven causal criteria were retained for further modeling. Results of validation using receiver operating characteristic (ROC) curves showed clear performance differences: GIS-MCDA with fuzzy standardization and AHP weighting achieved low predictive accuracy (area under the curve, AUC = 0.51), whereas MaxEnt performed strongly (AUC = 0.81). These findings highlight the ability of MaxEnt to capture nonlinear relationships and complex interactions among geoenvironmental causal criteria, whereas GIS-MCDA with fuzzy standardization and AHP proved inadequate for reliable wildfire susceptibility assessment in this context. These results confirm the findings of previous studies, while showing an even weaker performance of GIS-MCDA with fuzzy standardization and AHP compared to earlier studies. Importantly, this is the first application of machine learning through MaxEnt for wildfire susceptibility assessment in Montenegro, providing a spatial basis for wildfire management and a foundation for national-scale wildfire risk assessment.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 724-736"},"PeriodicalIF":4.1,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145624019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Temporal glacier velocity variations and their controlling factors in the Nathorstbreen glacier system, Svalbard 斯瓦尔巴群岛Nathorstbreen冰川系统冰川速度变化及其控制因素

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-12-01 DOI: 10.1016/j.ejrs.2025.11.004

Supratim Guha , Hyun-Cheol Kim

Temporal variations in glacier velocity are not only essential to understand glacier dynamics but also to predict glacier hazards. Therefore, in the current study, the continuous glacier velocities were estimated from 2014 to 2023 in the Nathorstbreen Glacier System (NGS), Svalbard, where a recent surge event has been observed. Also, the study identified and quantified the factors controlling variations of annual glacier velocity.

Using Landsat 8 OLI images, Cossi-corr (Co-registration of Optically Sensed Images and Correlation), an advanced Fourier-based image-matching tool, was utilized to estimate the velocity of the NGS. A multivariate regression analysis was performed to evaluate the influence of temperature, precipitation, snowfall, and terminus fluctuations on annual velocity changes.

The results indicate that the NGS exhibited the highest and lowest average annual velocities in 2021 and 2018, with magnitudes of 0.86 ± 0.11 m/day and 0.34 ± 0.18 m/day, respectively. The lower velocity in 2018 represents a quiescent phase following the previous surge, whereas the acceleration in 2021 reflects renewed dynamic activity linked to terminus retreat. Overall, glacier velocity declined from 2014 to 2018, increased between 2020 and 2022, and slightly decreased again in 2023. During this period, the glacier terminus experienced alternating annual retreat and advance, resulting in a net retreat of approximately 2.9 km. Terminus fluctuations were identified as important factors influencing annual glacier velocity, showing a lagged response between terminus movement and velocity. Including parameters such as ice thickness and subglacial hydrology in future analyses would further improve understanding of glacier velocity controls.

冰川流速的时间变化不仅是了解冰川动态的必要条件，而且也是预测冰川灾害的必要条件。因此，在本研究中，估计了2014年至2023年斯瓦尔巴群岛Nathorstbreen冰川系统（NGS）的连续冰川速度，在那里观测到最近的涌流事件。此外，研究还确定并量化了控制冰川年速度变化的因素。利用Landsat 8 OLI图像，利用基于傅里叶的先进图像匹配工具cossicorr （Co-registration of optical sensing images and Correlation）来估计NGS的速度。采用多元回归分析，评价了气温、降水、降雪量和终端线波动对年速度变化的影响。结果表明，NGS的年平均速度在2021年和2018年最高和最低，震级分别为0.86±0.11 m/d和0.34±0.18 m/d。2018年的较低速度代表了之前激增之后的静止阶段，而2021年的加速反映了与终端撤退相关的新的动态活动。总体而言，冰川速度在2014年至2018年期间下降，在2020年至2022年期间增加，并在2023年再次略有下降。在此期间，冰川末端经历了每年交替后退和前进的过程，导致净后退约2.9公里。终端波动是影响冰川年速度的重要因素，终端运动与速度之间存在滞后响应。在未来的分析中包括冰层厚度和冰下水文等参数将进一步提高对冰川速度控制的理解。

{"title":"Temporal glacier velocity variations and their controlling factors in the Nathorstbreen glacier system, Svalbard","authors":"Supratim Guha , Hyun-Cheol Kim","doi":"10.1016/j.ejrs.2025.11.004","DOIUrl":"10.1016/j.ejrs.2025.11.004","url":null,"abstract":"<div><div>Temporal variations in glacier velocity are not only essential to understand glacier dynamics but also to predict glacier hazards. Therefore, in the current study, the continuous glacier velocities were estimated from 2014 to 2023 in the Nathorstbreen Glacier System (NGS), Svalbard, where a recent surge event has been observed. Also, the study identified and quantified the factors controlling variations of annual glacier velocity.</div><div>Using Landsat 8 OLI images, Cossi-corr (Co-registration of Optically Sensed Images and Correlation), an advanced Fourier-based image-matching tool, was utilized to estimate the velocity of the NGS. A multivariate regression analysis was performed to evaluate the influence of temperature, precipitation, snowfall, and terminus fluctuations on annual velocity changes.</div><div>The results indicate that the NGS exhibited the highest and lowest average annual velocities in 2021 and 2018, with magnitudes of 0.86 ± 0.11 m/day and 0.34 ± 0.18 m/day, respectively. The lower velocity in 2018 represents a quiescent phase following the previous surge, whereas the acceleration in 2021 reflects renewed dynamic activity linked to terminus retreat. Overall, glacier velocity declined from 2014 to 2018, increased between 2020 and 2022, and slightly decreased again in 2023. During this period, the glacier terminus experienced alternating annual retreat and advance, resulting in a net retreat of approximately 2.9 km. Terminus fluctuations were identified as important factors influencing annual glacier velocity, showing a lagged response between terminus movement and velocity. Including parameters such as ice thickness and subglacial hydrology in future analyses would further improve understanding of glacier velocity controls.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 737-746"},"PeriodicalIF":4.1,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145693363","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Sentinel-2 and Planet-Scope as reliable tools for water quality monitoring of small reservoirs Sentinel-2和Planet-Scope是小型水库水质监测的可靠工具

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-11-17 DOI: 10.1016/j.ejrs.2025.11.002

Silvia Di Francesco , Francesco Biondi , Barbara Casentini , Stefano Fazi , Stefano Amalfitano , Marco D’Eugenio , Francesca Todisco , Stefano Casadei , Francesca Giannone

In an era marked by increasingly frequent extreme weather events, small inland reservoirs are emerging as crucial yet often overlooked water resources. This study investigates the potential of remote sensing techniques to efficiently monitor the water quality of those reservoirs and improve their management. Although many works in literature have tried to derive water quality parameters from different satellite platforms, micro satellites constellations like PlanetScope have never been investigated: they can be a promising tool for investigation of SRs thanks to their high spatial and temporal resolution. Focusing on Spina Reservoir, a small lake in the province of Perugia where a water quality survey has been conducted, the research combines on-site biochemical analyses with satellite imagery from Sentinel-2, well known and explored free Platform, and PlanetScope. The performances of images corrected with the default atmospheric correction and with a specific pre-processor (ACOLITE) for inland and coastal water are discussed.

Water Quality Semi-empirical algorithms (indices) based on one or more spectral bands at different wavelengths are used to build correlation curves respect to in-situ measurements (e.g Chlorophyll-a, turbidity, Cyanobacteria), enabling the evaluation and comparison of the performance.

PlanetScope images displayed higher reliability with respect to Sentinel-2 data and correction with ACOLITE lead to more accurate interpolations, except for chlorophyll-a, even if satellite images with lower spatial resolution (Sentinel-2) can also provide a well-distributed dataset.

The findings underscore the significant potential of PlanetScope microsatellite constellation for real-time, cost-effective water quality assessment that could be easily applied on a larger scale, as regional assessment.

在一个极端天气事件日益频繁的时代，小型内陆水库正成为至关重要但往往被忽视的水资源。本研究探讨了遥感技术在有效监测水库水质和改善水库管理方面的潜力。尽管文献中的许多作品都试图从不同的卫星平台上获得水质参数，但像PlanetScope这样的微型卫星星座从未被研究过：由于它们具有高时空分辨率，它们可以成为研究SRs的有前途的工具。该研究以佩鲁贾省的斯皮纳水库（Spina Reservoir）为研究对象，将现场生化分析与Sentinel-2卫星图像结合起来，Sentinel-2是著名的免费平台，PlanetScope也进行了水质调查。讨论了使用默认大气校正和特定预处理（ACOLITE）对内陆和沿海水域进行校正的图像的性能。基于不同波长的一个或多个光谱波段的半经验算法（指数）用于建立与原位测量（例如叶绿素-a，浊度，蓝藻）相关的曲线，从而对性能进行评估和比较。PlanetScope图像相对于Sentinel-2数据显示出更高的可靠性，即使空间分辨率较低的卫星图像（Sentinel-2）也可以提供分布良好的数据集，但ACOLITE校正导致更准确的插值（叶绿素-a除外）。这些发现强调了PlanetScope微卫星星座在实时、具有成本效益的水质评估方面的巨大潜力，可以很容易地应用于更大规模的区域评估。

{"title":"Sentinel-2 and Planet-Scope as reliable tools for water quality monitoring of small reservoirs","authors":"Silvia Di Francesco , Francesco Biondi , Barbara Casentini , Stefano Fazi , Stefano Amalfitano , Marco D’Eugenio , Francesca Todisco , Stefano Casadei , Francesca Giannone","doi":"10.1016/j.ejrs.2025.11.002","DOIUrl":"10.1016/j.ejrs.2025.11.002","url":null,"abstract":"<div><div>In an era marked by increasingly frequent extreme weather events, small inland reservoirs are emerging as crucial yet often overlooked water resources. This study investigates the potential of remote sensing techniques to efficiently monitor the water quality of those reservoirs and improve their management. Although many works in literature have tried to derive water quality parameters from different satellite platforms, micro satellites constellations like PlanetScope have never been investigated: they can be a promising tool for investigation of SRs thanks to their high spatial and temporal resolution. Focusing on Spina Reservoir, a small lake in the province of Perugia where a water quality survey has been conducted, the research combines on-site biochemical analyses with satellite imagery from Sentinel-2, well known and explored free Platform, and PlanetScope. The performances of images corrected with the default atmospheric correction and with a specific pre-processor (ACOLITE) for inland and coastal water are discussed.</div><div>Water Quality Semi-empirical algorithms (indices) based on one or more spectral bands at different wavelengths are used to build correlation curves respect to in-situ measurements (e.g Chlorophyll-a, turbidity, Cyanobacteria), enabling the evaluation and comparison of the performance.</div><div>PlanetScope images displayed higher reliability with respect to Sentinel-2 data and correction with ACOLITE lead to more accurate interpolations, except for chlorophyll-a, even if satellite images with lower spatial resolution (Sentinel-2) can also provide a well-distributed dataset.</div><div>The findings underscore the significant potential of PlanetScope microsatellite constellation for real-time, cost-effective water quality assessment that could be easily applied on a larger scale, as regional assessment.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 713-723"},"PeriodicalIF":4.1,"publicationDate":"2025-11-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145578964","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Iron ore exploration in the Central Eastern Desert of Egypt: Insights from remote Sensing, Geophysical, and geochemical data 埃及中东部沙漠的铁矿勘探：来自遥感、地球物理和地球化学数据的见解

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-11-11 DOI: 10.1016/j.ejrs.2025.11.001

Mahmoud Abd El-Rahman Hegab , Salem Mohamed Salem , Nehal Mohamed Soliman , Kareem Hamed Abd El Wahid , Soha Hassan , Alaa Nayef , Mohamed Anwar Ahmed

The novelty of this study lies in applying an integrated workflow that combines geological mapping, aeromagnetic analysis, remote sensing, and XRF analysis to delineate extensions of known iron ore deposits and identify previously unrecognized occurrences, while simultaneously providing new insights into the tectono-magmatic controls of iron mineralization in the Central Eastern Desert. The findings provide critical data on the spatial organization, mineral characteristics, and geological controls of iron ore in this complex tectonic setting, enabling more efficient exploration plans in the Eastern Desert. Metamorphosed banded iron formations (BIFs) prevail at several localities, e.g., Gabal El Hadid, Umm Nar, Umm Ghamis El Zarqa, El Sibai, El Dabbah, and Wadi Kareem. These BIFs occur within a metavolcano-sedimentary environment, with thicknesses of up to 5 m, in the form of bands and lenses composed of magnetite, hematite, and silica. Magnetic spectral analysis enabled clear discrimination among lithological units, definition of structural controls, and demarcation of alteration zones associated with iron mineralization.

这项研究的新颖之处在于，它将地质测绘、航磁分析、遥感和XRF分析相结合，应用了一个集成的工作流程，以描绘已知铁矿床的延伸，并识别以前未被识别的矿点，同时为中东部沙漠中铁矿化的构造-岩浆控制提供了新的见解。这一发现为研究这一复杂构造环境下的铁矿石空间组织、矿物特征和地质控制提供了关键数据，有助于制定更有效的东部沙漠勘探计划。变质带状铁地层（BIFs）在Gabal El Hadid、Umm Nar、Umm Ghamis El Zarqa、El Sibai、El Dabbah和Wadi Kareem等几个地方普遍存在。这些bif出现在变质火山-沉积环境中，厚度可达5米，以磁铁矿、赤铁矿和二氧化硅组成的带状和透镜状的形式存在。通过磁谱分析，可以清晰地区分岩性单元，明确构造控制，划分与铁矿成矿有关的蚀变带。

{"title":"Iron ore exploration in the Central Eastern Desert of Egypt: Insights from remote Sensing, Geophysical, and geochemical data","authors":"Mahmoud Abd El-Rahman Hegab , Salem Mohamed Salem , Nehal Mohamed Soliman , Kareem Hamed Abd El Wahid , Soha Hassan , Alaa Nayef , Mohamed Anwar Ahmed","doi":"10.1016/j.ejrs.2025.11.001","DOIUrl":"10.1016/j.ejrs.2025.11.001","url":null,"abstract":"<div><div>The novelty of this study lies in applying an integrated workflow that combines geological mapping, aeromagnetic analysis, remote sensing, and XRF analysis to delineate extensions of known iron ore deposits and identify previously unrecognized occurrences, while simultaneously providing new insights into the tectono-magmatic controls of iron mineralization in the Central Eastern Desert. The findings provide critical data on the spatial organization, mineral characteristics, and geological controls of iron ore in this complex tectonic setting, enabling more efficient exploration plans in the Eastern Desert. Metamorphosed banded iron formations (BIFs) prevail at several localities, e.g., Gabal El Hadid, Umm Nar, Umm Ghamis El Zarqa, El Sibai, El Dabbah, and Wadi Kareem. These BIFs occur within a metavolcano-sedimentary environment, with thicknesses of up to 5 m, in the form of bands and lenses composed of magnetite, hematite, and silica. Magnetic spectral analysis enabled clear discrimination among lithological units, definition of structural controls, and demarcation of alteration zones associated with iron mineralization.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 690-712"},"PeriodicalIF":4.1,"publicationDate":"2025-11-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145528367","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advancing coastal land use mapping through deep multi-label classification and multi-sensor data fusion 通过深度多标签分类和多传感器数据融合推进沿海土地利用制图

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-10-24 DOI: 10.1016/j.ejrs.2025.10.004

Alireza Sharifi , Mohammad Mahdi Safari , Bayan Alabdullah

Coastal environments change and are environmentally sensitive. Land use classification must be accurate and timely for sustainable development, environmental monitoring, and catastrophe risk management. This research introduces a deep learning framework for categorizing coastal land use with multiple labels using high-resolution satellite pictures from several sensors. We design and evaluate a deep convolutional neural network architecture that classifies photos with multiple labels optimally using the MLRSNet dataset, which comprises 60 semantic classes from Chinese coastal locations. Data fusion merges spectral, spatial, and textural characteristics from many remote sensing methods, making classification findings more trustworthy and relevant to more circumstances. Numerous studies have proven that our method accurately separates complex and visually similar coastal categories including wetlands, beaches, rivers, ships, and urban coastlines. Precision, recall, F1-score, and mAP are used to evaluate the model. We also analyze its performance and mistakes in each class. The results demonstrate how deep learning and data fusion may address coastal remote sensing issues such semantic ambiguity, class variability, and class imbalance. This study enhances geographic artificial intelligence (GeoAI) by showing how to create a high-resolution shoreline map using a framework that works from start to end, can be scaled up, and can be utilized elsewhere. The recommended strategy affects environmental monitoring, coastal zone management, and fact-based decision-making, notably with climate change and urbanization along the coastline. Deep learning and multi-sensor satellite technologies can improve operational coastal monitoring systems, according to our findings.

沿海环境不断变化，对环境十分敏感。土地利用分类必须准确、及时地用于可持续发展、环境监测和巨灾风险管理。本研究引入了一个深度学习框架，利用来自多个传感器的高分辨率卫星图片，对沿海土地使用进行多标签分类。我们设计并评估了一个深度卷积神经网络架构，该架构使用MLRSNet数据集对具有多个标签的照片进行最佳分类，该数据集包含来自中国沿海地区的60个语义类。数据融合融合了许多遥感方法的光谱、空间和纹理特征，使分类结果更加可信，适用于更多的情况。许多研究已经证明，我们的方法可以准确地分离复杂的和视觉上相似的海岸类别，包括湿地、海滩、河流、船舶和城市海岸线。使用Precision， recall， F1-score和mAP来评估模型。我们还分析了它在每节课上的表现和错误。研究结果表明，深度学习和数据融合可以解决沿海遥感问题，如语义模糊、类别可变性和类别不平衡。这项研究通过展示如何使用从头到尾工作的框架创建高分辨率海岸线地图来增强地理人工智能（GeoAI），该框架可以按比例放大，并可以在其他地方使用。建议的策略影响环境监测、海岸带管理和基于事实的决策，特别是沿海地区的气候变化和城市化。根据我们的研究结果，深度学习和多传感器卫星技术可以改善沿海监测系统的运行。

{"title":"Advancing coastal land use mapping through deep multi-label classification and multi-sensor data fusion","authors":"Alireza Sharifi , Mohammad Mahdi Safari , Bayan Alabdullah","doi":"10.1016/j.ejrs.2025.10.004","DOIUrl":"10.1016/j.ejrs.2025.10.004","url":null,"abstract":"<div><div>Coastal environments change and are environmentally sensitive. Land use classification must be accurate and timely for sustainable development, environmental monitoring, and catastrophe risk management. This research introduces a deep learning framework for categorizing coastal land use with multiple labels using high-resolution satellite pictures from several sensors. We design and evaluate a deep convolutional neural network architecture that classifies photos with multiple labels optimally using the MLRSNet dataset, which comprises 60 semantic classes from Chinese coastal locations. Data fusion merges spectral, spatial, and textural characteristics from many remote sensing methods, making classification findings more trustworthy and relevant to more circumstances. Numerous studies have proven that our method accurately separates complex and visually similar coastal categories including wetlands, beaches, rivers, ships, and urban coastlines. Precision, recall, F1-score, and mAP are used to evaluate the model. We also analyze its performance and mistakes in each class. The results demonstrate how deep learning and data fusion may address coastal remote sensing issues such semantic ambiguity, class variability, and class imbalance. This study enhances geographic artificial intelligence (GeoAI) by showing how to create a high-resolution shoreline map using a framework that works from start to end, can be scaled up, and can be utilized elsewhere. The recommended strategy affects environmental monitoring, coastal zone management, and fact-based decision-making, notably with climate change and urbanization along the coastline. Deep learning and multi-sensor satellite technologies can improve operational coastal monitoring systems, according to our findings.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 681-689"},"PeriodicalIF":4.1,"publicationDate":"2025-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145363017","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Deep multimodal unmixing of hyperspectral images using Convolutional Block Attention Module (CBAM) and LiDAR features 基于卷积块注意模块（CBAM）和激光雷达特征的高光谱图像深度多模态解混

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-10-21 DOI: 10.1016/j.ejrs.2025.10.002

M Sreejam, L Agilandeeswari

Hyperspectral image unmixing has garnered considerable attention across various application domains, particularly remote sensing applications. However, relying solely on one modality to distinguish objects with similar spectral information presents several shortcomings. Enhanced performance can be achieved by integrating geographical information from Light Detection and Ranging (LiDAR) data into Unmixing. This paper introduces a new unmixing model that combines hyperspectral and LiDAR data. Impressive data representation and feature extraction using deep learning technology have been employed to develop the Multimodal Hyperspectral Unmixing Model using CBAM (Convolutional Block Attention Module) attention (MHUCBAM). The model exemplifies a sophisticated approach to multimodal unmixing, incorporating Spectral Spatial attention alongside the CBAM. Channel Attention improved the model’s capability to analyze complex spatial and spectral relationships. Our model achieves accurate unmixing of complex environments with effective multimodal data representation and deep feature extraction. Two real-world multimodal unmixing datasets, namely, Houston and Muffle, are used for the performance evaluation. A rigorous ablation analysis was performed to validate the performance of the proposed model. The comparative study with existing unmixing models demonstrated that utilizing latent features from LiDAR data resulted in better unmixing outcomes in terms of both Root Mean Square Error (RMSE) and Spectral Angular Distance (SAD).

高光谱图像解混已经在各个应用领域引起了相当大的关注，特别是遥感应用。然而，仅仅依靠一种模态来区分具有相似光谱信息的物体存在一些缺点。通过将来自光探测和测距（LiDAR）数据的地理信息集成到Unmixing中，可以提高性能。本文介绍了一种结合高光谱和激光雷达数据的新解混模型。使用深度学习技术的令人印象深刻的数据表示和特征提取被用于开发使用CBAM（卷积块注意模块）注意（MHUCBAM）的多模态高光谱解混模型。该模型体现了一种复杂的多模态解混方法，将频谱空间关注与CBAM结合在一起。通道注意提高了模型分析复杂空间和光谱关系的能力。该模型通过有效的多模态数据表示和深度特征提取实现了复杂环境的精确解混。两个真实世界的多模态解混数据集，即Houston和Muffle，用于性能评估。进行了严格的烧蚀分析，以验证所提出模型的性能。与现有解混模型的对比研究表明，利用LiDAR数据的潜在特征在均方根误差（RMSE）和光谱角距离（SAD）方面都可以获得更好的解混结果。

{"title":"Deep multimodal unmixing of hyperspectral images using Convolutional Block Attention Module (CBAM) and LiDAR features","authors":"M Sreejam, L Agilandeeswari","doi":"10.1016/j.ejrs.2025.10.002","DOIUrl":"10.1016/j.ejrs.2025.10.002","url":null,"abstract":"<div><div>Hyperspectral image unmixing has garnered considerable attention across various application domains, particularly remote sensing applications. However, relying solely on one modality to distinguish objects with similar spectral information presents several shortcomings. Enhanced performance can be achieved by integrating geographical information from Light Detection and Ranging (LiDAR) data into Unmixing. This paper introduces a new unmixing model that combines hyperspectral and LiDAR data. Impressive data representation and feature extraction using deep learning technology have been employed to develop the Multimodal Hyperspectral Unmixing Model using CBAM (Convolutional Block Attention Module) attention (MHUCBAM). The model exemplifies a sophisticated approach to multimodal unmixing, incorporating Spectral Spatial attention alongside the CBAM. Channel Attention improved the model’s capability to analyze complex spatial and spectral relationships. Our model achieves accurate unmixing of complex environments with effective multimodal data representation and deep feature extraction. Two real-world multimodal unmixing datasets, namely, Houston and Muffle, are used for the performance evaluation. A rigorous ablation analysis was performed to validate the performance of the proposed model. The comparative study with existing unmixing models demonstrated that utilizing latent features from LiDAR data resulted in better unmixing outcomes in terms of both Root Mean Square Error (RMSE) and Spectral Angular Distance (SAD).</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 666-680"},"PeriodicalIF":4.1,"publicationDate":"2025-10-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145363018","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Efficient monitoring of groundwater level changes using compressive remote sensing 压缩遥感对地下水位变化的有效监测

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-10-17 DOI: 10.1016/j.ejrs.2025.10.003

Cihan Bayındır , Ali Rıza Alan

In this paper, we propose and discuss the applicability of compressive sensing (CS) for the remote measurement and analysis of groundwater level changes. For this purpose, we consider three watersheds in Turkey and utilize the data acquired by the Gravity Recovery and Climate Experiment (GRACE) satellite at these watersheds. These watersheds are Fırat (Euphrates), Kızılırmak, and Büyük Menderes (Greater Menderes). The data collected by the GRACE satellite have a temporal resolution on the order of months, however, due to operation and maintenance considerations it is known that some of the GRACE data may be missing. Using the time series data collected between 2002 and 2019 at these three watersheds we show that the time series of the groundwater table (GWT) can be reconstructed using CS which utilizes fewer samples than the classical Shannon’s theorem states. Thus, when the CS technique is utilized, measurement times and hardware storage requirements of groundwater sensing systems can be significantly reduced where some errors can be observed in the reconstruction of the GWT level. In some cases, such parameters can be exactly reconstructed by CS even in the presence of missing data if certain sparsity and sampling conditions are satisfied. The CS-based GWT reconstruction technique proposed in this paper can also be extended to measure and analyze other types of data such as in situ groundwater levels, groundwater velocities, and groundwater volume flux data in hydrology and hydraulics.

本文提出并讨论了压缩感知（CS）技术在地下水位变化遥感测量与分析中的适用性。为此，我们考虑了土耳其的三个流域，并利用了重力恢复和气候实验（GRACE）卫星在这些流域获得的数据。这些流域分别是Fırat（幼发拉底河）、Kızılırmak和b y k Menderes（大Menderes）。GRACE卫星收集的数据具有月级的时间分辨率，但是，由于操作和维护方面的考虑，已知一些GRACE数据可能会丢失。利用2002年至2019年在这三个流域收集的时间序列数据，我们表明使用CS可以重建地下水位（GWT）的时间序列，该方法比经典香农定理使用的样本更少。因此，当使用CS技术时，地下水传感系统的测量次数和硬件存储要求可以大大减少，但在重建GWT水位时可以观察到一些误差。在某些情况下，如果满足一定的稀疏性和采样条件，即使存在缺失数据，CS也可以精确地重建这些参数。本文提出的基于cs的GWT重建技术还可以扩展到其他类型的数据，如水文水力学中的地下水位、地下流速、地下水体积通量等数据的测量和分析。

{"title":"Efficient monitoring of groundwater level changes using compressive remote sensing","authors":"Cihan Bayındır , Ali Rıza Alan","doi":"10.1016/j.ejrs.2025.10.003","DOIUrl":"10.1016/j.ejrs.2025.10.003","url":null,"abstract":"<div><div>In this paper, we propose and discuss the applicability of compressive sensing (CS) for the remote measurement and analysis of groundwater level changes. For this purpose, we consider three watersheds in Turkey and utilize the data acquired by the Gravity Recovery and Climate Experiment (GRACE) satellite at these watersheds. These watersheds are Fırat (Euphrates), Kızılırmak, and Büyük Menderes (Greater Menderes). The data collected by the GRACE satellite have a temporal resolution on the order of months, however, due to operation and maintenance considerations it is known that some of the GRACE data may be missing. Using the time series data collected between 2002 and 2019 at these three watersheds we show that the time series of the groundwater table (GWT) can be reconstructed using CS which utilizes fewer samples than the classical Shannon’s theorem states. Thus, when the CS technique is utilized, measurement times and hardware storage requirements of groundwater sensing systems can be significantly reduced where some errors can be observed in the reconstruction of the GWT level. In some cases, such parameters can be exactly reconstructed by CS even in the presence of missing data if certain sparsity and sampling conditions are satisfied. The CS-based GWT reconstruction technique proposed in this paper can also be extended to measure and analyze other types of data such as in situ groundwater levels, groundwater velocities, and groundwater volume flux data in hydrology and hydraulics.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 659-665"},"PeriodicalIF":4.1,"publicationDate":"2025-10-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145325283","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Advanced time series forecasting of vegetation health using deep learning models: A remote sensing approach to analyzing climate change impact 使用深度学习模型的植被健康高级时间序列预测：一种分析气候变化影响的遥感方法

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-10-16 DOI: 10.1016/j.ejrs.2025.09.005

Sarhad Baez Hasan , Shahab Wahhab Kareem

The growing consequences of climate change on vegetation ecosystems require advanced predictive tools for environmental monitoring and adaptive management. This research explored a new application of hybrid deep learning models to forecast the Normalized Difference Vegetation Index (NDVI) time series, using Sentinel-2 high-resolution satellite images. Specifically, this research investigated vegetation dynamics in four climatically different regions of Northern Iraq from 2016 to 2024, developing and comparing eight deep learning models, including traditional recurrent networks (Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BiLSTM), and Gated Recurrent Unit (GRU)) and Convolutional Neural Networks (CNN), resulting in unique hybrid models that combine spatial and temporal feature extraction mechanisms. The study utilized a large dataset of 43,200 images with a spatial resolution of 10 m, employing systematic data preparation that included NDVI processing (NDVI calculations, normalization, and time-series sequence construction) necessary for model training and learning. The model performance was rigorously evaluated, where hybrid models were demonstrated to outperform other models, with BiLSTM-GRU appearing to deliver high accuracy (coefficient of determination scores R² of up to 0.851) and low prediction errors (Mean Squared Error (MSE) as low as 6.04 × 10⁻⁴). In terms of ecological region, model performance was assessed across regions, as well as across different regions, finding general trends in performance, particularly in regions with homogeneous vegetation cover at each time sampling period. The Monte Carlo dropout method offered the opportunity to infer uncertainty, which in turn helped build confidence in predictions. The predictions for the future periods of 2025–2028 show promising seasonal patterns and long-term trends, which are important with respect to climate-adjusted planning.

气候变化对植被生态系统的影响日益严重，需要先进的预测工具来进行环境监测和适应性管理。本研究探索了混合深度学习模型的新应用，利用Sentinel-2高分辨率卫星图像预测归一化植被指数（NDVI）时间序列。具体而言，本研究以2016 - 2024年伊拉克北部4个气候不同地区的植被动态为研究对象，开发并比较了8种深度学习模型，包括传统递归网络（长短期记忆（LSTM）、双向长短期记忆（BiLSTM）和门控递归单元（GRU））和卷积神经网络（CNN），形成了独特的结合时空特征提取机制的混合模型。本研究利用空间分辨率为10 m的43,200幅图像的大型数据集，采用系统的数据准备，包括NDVI处理（NDVI计算、归一化和时间序列序列构建），这是模型训练和学习所必需的。对模型的性能进行了严格的评估，混合模型被证明优于其他模型，BiLSTM-GRU似乎提供了高精度（决定系数R2高达0.851）和低预测误差（均方误差（MSE）低至6.04 × 10−4）。就生态区域而言，对模型的性能进行了跨区域和不同区域的评估，发现了性能的总体趋势，特别是在每个采样期植被覆盖均匀的区域。蒙特卡洛辍学法提供了推断不确定性的机会，这反过来又有助于建立对预测的信心。对2025-2028年未来时期的预测显示出有希望的季节模式和长期趋势，这对气候调整规划很重要。

{"title":"Advanced time series forecasting of vegetation health using deep learning models: A remote sensing approach to analyzing climate change impact","authors":"Sarhad Baez Hasan , Shahab Wahhab Kareem","doi":"10.1016/j.ejrs.2025.09.005","DOIUrl":"10.1016/j.ejrs.2025.09.005","url":null,"abstract":"<div><div>The growing consequences of climate change on vegetation ecosystems require advanced predictive tools for environmental monitoring and adaptive management. This research explored a new application of hybrid deep learning models to forecast the Normalized Difference Vegetation Index (NDVI) time series, using Sentinel-2 high-resolution satellite images. Specifically, this research investigated vegetation dynamics in four climatically different regions of Northern Iraq from 2016 to 2024, developing and comparing eight deep learning models, including traditional recurrent networks (Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (BiLSTM), and Gated Recurrent Unit (GRU)) and Convolutional Neural Networks (CNN), resulting in unique hybrid models that combine spatial and temporal feature extraction mechanisms. The study utilized a large dataset of 43,200 images with a spatial resolution of 10 m, employing systematic data preparation that included NDVI processing (NDVI calculations, normalization, and time-series sequence construction) necessary for model training and learning. The model performance was rigorously evaluated, where hybrid models were demonstrated to outperform other models, with BiLSTM-GRU appearing to deliver high accuracy (coefficient of determination scores R<sup>2</sup> of up to 0.851) and low prediction errors (Mean Squared Error (MSE) as low as 6.04 × 10<sup>−4</sup>). In terms of ecological region, model performance was assessed across regions, as well as across different regions, finding general trends in performance, particularly in regions with homogeneous vegetation cover at each time sampling period. The Monte Carlo dropout method offered the opportunity to infer uncertainty, which in turn helped build confidence in predictions. The predictions for the future periods of 2025–2028 show promising seasonal patterns and long-term trends, which are important with respect to climate-adjusted planning.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 645-658"},"PeriodicalIF":4.1,"publicationDate":"2025-10-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145325284","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

SpecSpatMamba: an efficient hyperspectral image classification method integrating spectral-spatial dual-path and state space model SpecSpatMamba：一种结合光谱空间双路径和状态空间模型的高效高光谱图像分类方法

IF 4.1 3区地球科学 Q2 ENVIRONMENTAL SCIENCES

Egyptian Journal of Remote Sensing and Space Sciences

Pub Date : 2025-10-08 DOI: 10.1016/j.ejrs.2025.10.001

Jianshang Liao , Liguo Wang

Current hyperspectral image classification methods face three critical limitations: (1) traditional CNNs suffer from the curse of dimensionality when processing high-dimensional spectral data, leading to overfitting and poor generalization; (2) existing approaches fail to effectively address spectral band redundancy, resulting in computational inefficiency and suboptimal feature representation; (3) conventional methods lack synergistic utilization of spatial-spectral information, treating spectral and spatial dimensions uniformly rather than exploiting their distinct characteristics. To address these gaps, this paper proposes SpecSpatMamba, a novel hyperspectral image classification method integrating spectral-spatial dual-path feature extraction with state space models. SpecSpatMamba introduces three core innovations: (1) Dual-path feature extraction with spectral-spatial separation, where 1 × 1 convolutions extract spectral features and 3 × 3 convolutions capture spatial features; (2) Hybrid architecture combining state space models with convolutional operations for balanced long-range dependency and local feature capture; (3) Computational efficiency breakthrough achieving O(L·d) linear complexity compared to Transformer’s O(L2·d) complexity. Experiments on four benchmark datasets—Indian Pines, Pavia University, Salinas Valley, and Houston2013—demonstrate competitive performance compared to state-of-the-art methods. SpecSpatMamba achieves overall accuracies of 95.11 %, 98.61 %, 96.97 %, and 91.48 %, respectively. Notably, SpecSpatMamba demonstrates superior cross-dataset consistency and robust performance across diverse geographic environments, with particularly strong improvements in complex urban scenarios (+0.39 % on Houston2013) and agricultural settings (+0.57 % on Salinas Valley), confirming the method’s effectiveness in addressing high-dimensional hyperspectral data challenges.

目前的高光谱图像分类方法面临三个关键的局限性：(1)传统cnn在处理高维光谱数据时存在维数诅咒，导致过拟合和泛化差；(2)现有方法无法有效处理频谱冗余，导致计算效率低下，特征表示不理想；(3)传统方法缺乏对空间光谱信息的协同利用，对光谱和空间维度进行统一处理，未能充分挖掘其各自的特征。为了解决这些问题，本文提出了SpecSpatMamba，一种将光谱-空间双路径特征提取与状态空间模型相结合的新型高光谱图像分类方法。SpecSpatMamba引入了三个核心创新：(1)光谱-空间分离双路径特征提取，其中1 × 1卷积提取光谱特征，3 × 3卷积捕获空间特征；(2)结合状态空间模型和卷积运算的混合架构，平衡远程依赖和局部特征捕获；(3)与Transformer的O（L2·d）复杂度相比，计算效率突破，实现了O（L·d）线性复杂度。在四个基准数据集（indian Pines、Pavia University、Salinas Valley和houston 2013）上进行的实验表明，与最先进的方法相比，它们的性能具有竞争力。SpecSpatMamba的总体准确率分别为95.11%、98.61%、96.97%和91.48%。值得注意的是，SpecSpatMamba在不同地理环境中表现出卓越的跨数据集一致性和稳健的性能，在复杂的城市场景（休斯顿2013年+ 0.39%）和农业环境（萨利纳斯山谷+ 0.57%）中表现出特别强的改进，证实了该方法在解决高维高光谱数据挑战方面的有效性。

{"title":"SpecSpatMamba: an efficient hyperspectral image classification method integrating spectral-spatial dual-path and state space model","authors":"Jianshang Liao , Liguo Wang","doi":"10.1016/j.ejrs.2025.10.001","DOIUrl":"10.1016/j.ejrs.2025.10.001","url":null,"abstract":"<div><div>Current hyperspectral image classification methods face three critical limitations: (1) traditional CNNs suffer from the curse of dimensionality when processing high-dimensional spectral data, leading to overfitting and poor generalization; (2) existing approaches fail to effectively address spectral band redundancy, resulting in computational inefficiency and suboptimal feature representation; (3) conventional methods lack synergistic utilization of spatial-spectral information, treating spectral and spatial dimensions uniformly rather than exploiting their distinct characteristics. To address these gaps, this paper proposes SpecSpatMamba, a novel hyperspectral image classification method integrating spectral-spatial dual-path feature extraction with state space models. SpecSpatMamba introduces three core innovations: (1) Dual-path feature extraction with spectral-spatial separation, where 1 × 1 convolutions extract spectral features and 3 × 3 convolutions capture spatial features; (2) Hybrid architecture combining state space models with convolutional operations for balanced long-range dependency and local feature capture; (3) Computational efficiency breakthrough achieving O(L·d) linear complexity compared to Transformer’s O(L2·d) complexity. Experiments on four benchmark datasets—Indian Pines, Pavia University, Salinas Valley, and Houston2013—demonstrate competitive performance compared to state-of-the-art methods. SpecSpatMamba achieves overall accuracies of 95.11 %, 98.61 %, 96.97 %, and 91.48 %, respectively. Notably, SpecSpatMamba demonstrates superior cross-dataset consistency and robust performance across diverse geographic environments, with particularly strong improvements in complex urban scenarios (+0.39 % on Houston2013) and agricultural settings (+0.57 % on Salinas Valley), confirming the method’s effectiveness in addressing high-dimensional hyperspectral data challenges.</div></div>","PeriodicalId":48539,"journal":{"name":"Egyptian Journal of Remote Sensing and Space Sciences","volume":"28 4","pages":"Pages 628-644"},"PeriodicalIF":4.1,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145269670","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0