Dual-Stream Feature Collaboration Perception Network for Salient Object Detection in Remote Sensing Images DOI Open Access
Hongli Li,

Xuhui Chen,

Liye Mei

и другие.

Electronics, Год журнала: 2024, Номер 13(18), С. 3755 - 3755

Опубликована: Сен. 21, 2024

As the core technology of artificial intelligence, salient object detection (SOD) is an important approach to improve analysis efficiency remote sensing images by intelligently identifying key areas in images. However, existing methods that rely on a single strategy, convolution or Transformer, exhibit certain limitations complex scenarios. Therefore, we developed Dual-Stream Feature Collaboration Perception Network (DCPNet) enable collaborative work and feature complementation Transformer CNN. First, adopted dual-branch extractor with strong local bias long-range dependence characteristics perform multi-scale extraction from Then, presented Multi-path Complementary-aware Interaction Module (MCIM) refine fuse representations targets global branches, achieving fine-grained fusion interactive alignment features. Finally, proposed Weighting Balance (FWBM) balance features, preventing model overemphasizing information at expense details inadequately mining cues due excessive focus information. Extensive experiments EORSSD ORSSD datasets demonstrated DCPNet outperformed current 19 state-of-the-art methods.

Язык: Английский

Urban wetland landscape patterns and cooling effects in Guilin utilizing GF-1/6 and SDGSAT-1 data DOI Creative Commons
Ziqi Meng, Huadong Guo, Jingjuan Liao

и другие.

International Journal of Digital Earth, Год журнала: 2025, Номер 18(1)

Опубликована: Фев. 24, 2025

Язык: Английский

Процитировано

0
Research Trends in Vegetation Spatiotemporal Dynamics and Driving Forces: A Bibliometric Analysis (1987–2024) DOI Open Access

Dejin Dong,

Jianbo Shen, Daohong Gong

и другие.

Forests, Год журнала: 2025, Номер 16(4), С. 588 - 588

Опубликована: Март 28, 2025

Under the dual pressures of climate change and rapid urbanization, a comprehensive analysis vegetation’s spatiotemporal patterns their driving forces plays pivotal role for addressing global ecological challenges. However, systematic bibliometric analyses in this field remain limited. This study involved 18,270 related publications from 1989 to 2024 retrieved Web Science SCI-Expanded database, elucidating research trends, methodologies, key thematic areas. Utilizing bibliometrix biblioshiny tools, results reveal an annual average growth rate 17.62% number published articles, indicating expansion. Climate emerged as core force, with high-frequency keywords such “vegetation”, “dynamics”, “variability”. China (18,687 papers), United States (14,502 Germany (3394 papers) are leading contributors domain, showing fastest output, albeit relatively lower citation rates. Core journals, including Remote Sensing Environment Global Change Biology, have played roles advancing vegetation dynamics research, remote sensing techniques dominating field. The highlights shift single-variable (e.g., temperature, precipitation) multi-scale multidimensional approaches around 2010. Regional studies, those focusing on Loess Plateau, gaining importance, while advancements machine learning technologies enhanced precision scalability research. provides summary current state development trends forces, offering valuable insights future

Язык: Английский

Процитировано

0
Mitigation of human activity impacts on habitat quality in the Chengdu–Chongqing urban agglomeration DOI Creative Commons

Long Wan,

Ling Long,

Ping Xie

и другие.

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 16, 2025

Язык: Английский

Процитировано

0
Spatiotemporal Evolution Characteristics and Prediction of Habitat Quality Changes in the Poyang Lake Region, China DOI Open Access
Yu Liu,

Junxin Zhou,

Chenggong Liu

и другие.

Sustainability, Год журнала: 2025, Номер 17(8), С. 3708 - 3708

Опубликована: Апрель 19, 2025

The terrestrial spatial patterns were affected by human activities, primarily on regional land use (LU) changes, with habitat quality (HQ) serving as a prerequisite for achieving sustainable development. Assessing and predicting the spatiotemporal evolution characteristics of LU changes HQ is critical formulating strategies enhancing ecosystem service functions. Using Poyang Lake Region our research object, this employs data utilizes ‘InVEST’ model hot-spot analysis to quantitatively evaluate in during 2000–2020. PLUS then applied predict trends from 2020 2050. findings are follows: (1). From 2000 2020, areas forestland, shrubland, sparse woodland, paddy fields, dryland showed decreasing trend, reductions mainly occurring urban expansion zones such Nanchang City largely converted into construction land. (2). Since 2000, has shown slight retrogressive evolution, significant heterogeneity. spatially exhibits pattern improvement radiating outward major cities. (3). Predictions 2030 2050 indicate that will continue decline, most downward built-up their peripheries. reveal an ring around east–west corridor linking Pingxiang, Yichun, Xinyu, Nanchang, Fuzhou, Yingtan, Shangrao. This study provided basis direction planning policies its surrounding areas, while also contributing agrarian security enhancement levels region.

Язык: Английский

Процитировано

0
Identifying Regulatory Barriers in the Management of Ecological Corridors in an Increasingly Congested Space DOI

Orit Rotem,

Oren Perez,

N. Katzir

и другие.

SSRN Electronic Journal, Год журнала: 2025, Номер unknown

Опубликована: Янв. 1, 2025

Язык: Английский

Процитировано

0
Machine Learning for Criteria Weighting in GIS-Based Multi-Criteria Evaluation: A Case Study of Urban Suitability Analysis DOI Creative Commons

Lan Qing Zhao,

Alysha van Duynhoven, Suzana Dragićević

и другие.

Land, Год журнала: 2024, Номер 13(8), С. 1288 - 1288

Опубликована: Авг. 15, 2024

Geographic Information System-based Multi-Criteria Evaluation (GIS-MCE) methods are designed to assist in various spatial decision-making problems using data. Deriving criteria weights is an important component of GIS-MCE, typically relying on stakeholders’ opinions or mathematical methods. These approaches can be costly, time-consuming, and prone subjectivity bias. Therefore, the main objective this study investigate use Machine Learning (ML) techniques support weight derivation within GIS-MCE. The proposed ML-MCE method explored a case urban development suitability analysis City Kelowna, Canada. Feature importance values drawn from three ML techniques–Random Forest (RF), Extreme Gradient Boosting (XGB), Support Vector (SVM)–are used derive weights. scores obtained methodology compared with Equal-Weights (EW) Analytical Hierarchy Process (AHP) approach for weighting. results indicate that ML-derived where RF XGB provide more similar than those derived SVM. similarities differences confirmed Kappa indices comparing pairs maps. new processes land-use planning.

Язык: Английский

Процитировано

2
DCFF-Net: Deep Context Feature Fusion Network for High-Precision Classification of Hyperspectral Image DOI Creative Commons

Zhoupeng Chen,

Yu Chen, Yuan Wang

и другие.

Remote Sensing, Год журнала: 2024, Номер 16(16), С. 3002 - 3002

Опубликована: Авг. 15, 2024

Hyperspectral images (HSI) contain abundant spectral information. Efficient extraction and utilization of this information for image classification remain prominent research topics. Previously, hyperspectral techniques primarily relied on statistical attributes mathematical models data. Deep learning have recently been extensively utilized data classification, yielding promising outcomes. This study proposes a deep approach that uses polarization feature maps classification. Initially, the polar co-ordinate transformation method was employed to convert all pixels in into maps. Subsequently, proposed Context Feature Fusion Network (DCFF-NET) classify these The model validated using three open-source datasets: Indian Pines, Pavia University, Salinas. experimental results indicated DCFF-NET achieved excellent performance. Experimental public HSI datasets demonstrated accurately recognized different objects with an overall accuracy (OA) 86.68%, 94.73%, 95.14% based pixel method, 98.15%, 99.86%, 99.98% pixel-patch method.

Язык: Английский

Процитировано

2
Analysis of Spatiotemporal Evolution and Driving Forces of Vegetation from 2001 to 2020: A Case Study of Shandong Province, China DOI Open Access

Dejin Dong,

Ziliang Zhao,

Hongdi Gao

и другие.

Forests, Год журнала: 2024, Номер 15(7), С. 1245 - 1245

Опубликована: Июль 17, 2024

As global climate change intensifies and human activities escalate, changes in vegetation cover, an important ecological indicator, hold significant implications for ecosystem protection management. Shandong Province, a critical agricultural economic zone China, experiences that crucially affect regional regulation biodiversity conservation. This study employed normalized difference index (NDVI) data, combined with climatic, topographic, anthropogenic activity utilizing trend analysis methods, partial correlation analysis, Geodetector to comprehensively analyze the spatiotemporal variations primary driving factors of cover Province from 2001 2020. The findings indicate overall upward particularly areas concentrated activities. Climatic factors, such as precipitation temperature, exhibit positive growth, while land use emerge one key drivers influencing dynamics. Additionally, topography also impacts spatial distribution certain extent. research provides scientific basis management similar regions, supporting formulation effective restoration conservation strategies.

Язык: Английский

Процитировано

2
Spatiotemporal Dynamics and Prediction of Habitat Quality Based on Land Use and Cover Change in Jiangsu, China DOI Creative Commons
Ge Shi, Chuang Chen,

Qianqian Cao

и другие.

Remote Sensing, Год журнала: 2024, Номер 16(22), С. 4158 - 4158

Опубликована: Ноя. 7, 2024

Analyzing the spatiotemporal evolution characteristics of urban land use and habitat quality is crucial for sustainable development ecological environments. This study utilizes data Jiangsu Province years 2000, 2010, 2020, applying FLUS model to investigate driving force behind expansion simulate a prediction 2030. By integrating InVEST landscape pattern indices, this analyzes in uses geographical detector analysis examine synergistic effects influencing factors. The results indicate that, from 2000 degradation progressively increased, with spatial distribution levels showing gradual change. Under protection scenario 2030, fragmentation was alleviated. Conversely, under economic scenario, further deteriorated, resulting largest area low-quality regions. Minimal changes occurred natural scenario. (2) indices experienced significant 2020. continuous into other types led trend fragmentation, clear increasing dispersion, sprawl, Shannon’s diversity index, accompanied by decrease cohesion. (3) dominant interacting factors affecting were combinations socioeconomic factors, indicating that economy largely determines quality. findings provide optimization strategies future planning offer references restoration efforts region.

Язык: Английский

Процитировано

2
Ecological Security Pattern Construction and Multi-Scenario Risk Early Warning (2020–2035) in the Guangdong–Hong Kong–Macao Greater Bay Area, China DOI Creative Commons
Junjie Ma, Zhixiong Mei, Xinyu Wang

и другие.

Land, Год журнала: 2024, Номер 13(8), С. 1267 - 1267

Опубликована: Авг. 12, 2024

The effectiveness of ecological security patterns (ESPs) in maintaining regional stability and promoting sustainable development is widely recognized. However, limited research has focused on the early warning risks inherent ESPs. In this study, Guangdong–Hong Kong–Macao Greater Bay Area (GHKMGBA) taken as study area, risk zones are delineated by combining landscape index habitat quality, a multi-level ESP constructed based circuit theory. PLUS model was employed to simulate future built-up land expansion under different scenarios, which were then extracted overlaid with enable multi-scenario risks. results showed following: central plains coastal areas GHKMGBA exhibits high level risk, whereas peripheral forested face less threat, crucial for stability. ESP, comprising sources, corridors, pinch points, flow stability, tertiary corridors significant stress all requiring restoration enhancement efforts. There differences severity within across various scenarios. Under protection scenario, will have best situation, effectively protecting reducing damage, providing valuable reference policies. it must not overlook economic still needs further seek balance between growth protection.

Язык: Английский

Процитировано

1