Enhancing Land Cover Classification via Deep Ensemble Network

Knowledge-Based Systems, Journal Year: 2024, Volume and Issue: 305, P. 112611 - 112611

Published: Oct. 11, 2024

Language: Английский

---

PERSPECTIVAS BIBLIOMÉTRICAS SOBRE O SENSORIAMENTO REMOTO NA CLASSIFICAÇÃO DO SOLO URBANO

Revista Políticas Públicas & Cidades, Journal Year: 2025, Volume and Issue: 14(1), P. e1292 - e1292

Published: Jan. 16, 2025

Esta pesquisa examina detalhadamente a evolução das metodologias e tecnologias utilizadas na classificação do uso cobertura solo em áreas urbanas, com ênfase aplicação sensoriamento remoto. Por meio de uma análise bibliométrica sistemática, investigamos as tendências publicação inovações metodológicas que têm surgido no campo. Identificamos padrões significativos publicações, destacando autores influentes, instituições periódicos contribuem para o avanço conhecimento nessa área. Utilizamos método ProKnow-C, métricas como índice-h contagem citações avaliar influência impacto dos trabalhos científicos, oferecendo compreensão da dinâmica acadêmica redes colaboração existentes. Além disso, discutimos os desafios atuais enfrentados urbano novas podem ser aplicadas superá-los. Propomos direções promissoras pesquisas futuras, enfatizando importância desenvolvimento modelos sistemas monitoramento adaptativos inteligentes. Esses são fundamentais promover gestão urbana sustentável, permitindo cidades se adaptem eficazmente às mudanças ambientais, socioeconômicas demandas crescentes população.

---

Design and simulation of an intelligent irrigation system using fuzzy logic

Deleted Journal, Journal Year: 2025, Volume and Issue: 2(1)

Published: Jan. 24, 2025

Language: Английский

---

Improving remote sensing dehazing quality through local hybrid correction and optimization of atmospheric attenuation model based on wavelength

Frontiers in Remote Sensing, Journal Year: 2025, Volume and Issue: 5

Published: Jan. 30, 2025

Near-ground remote sensing image dehazing is crucial for accurately monitoring land resources. An effective technique and a precise atmospheric attenuation model are fundamental to acquiring real-time ground data with high fidelity. The dark channel prior (DCP) widely used method improving visibility in hazy conditions, but it often results reduced clarity artifacts, that limit its practical utility. To address these limitations, we propose novel hybrid correction method, local (LHC), which integrates gamma high-contrast regions logarithmic low-contrast within dehazed image. We calculated the cumulative distribution function (CDF) of Weber contrast analyzed impact different thresholds on effectiveness reducing artifacts. Our showed threshold corresponding 90% CDF significantly improved sharpness artifacts compared other thresholds. Furthermore, LHC outperformed both corrections terms artifact reduction, even after applying additional post-processing methods such as multi-exposure fusion guided filtering. quantitative analysis images, using gray-level co-occurrence matrix (GLCM) metrics, indicated offered balanced advantage enhancing details, texture consistency, structural complexity. Specifically, images processed by exhibit moderate correlation, low homogeneity entropy, all made very suitable solution near-ground tasks required enhanced detail also examined coefficient, observing increased distance, deviating progressively from empirical values, this phenomenon underscored complex effects scattering accuracy, especially at extended ranges. Additionally, refined transmittance light reflection 550 nm wavelength verdant landscapes, model’s alignment real-world conditions. This approach was not only could adapt wavelengths future studies. Overall, our research advanced precision techniques, promising decision-making resource management variety environmental applications.

Language: Английский

---

MobileTransNeXt: Integrating CNN, transformer, and BiLSTM for image classification

Alexandria Engineering Journal, Journal Year: 2025, Volume and Issue: 123, P. 460 - 470

Published: March 28, 2025

Language: Английский

---

A Novel Active Learning Approach for Improving Classification of Unlabeled Video Based on Deep Learning Techniques

Journal of Social Computing, Journal Year: 2025, Volume and Issue: 6(1), P. 1 - 17

Published: March 1, 2025

Language: Английский

---

DLAN: A Dual Attention Network for Effective Land Cover Classification in Remote Sensing

Knowledge-Based Systems, Journal Year: 2025, Volume and Issue: unknown, P. 113620 - 113620

Published: April 1, 2025

Language: Английский

---

C2-Net: Improving Feature Extraction and Alignment for Few-Shot Fine-Grained Image Classification

Research Square (Research Square), Journal Year: 2025, Volume and Issue: unknown

Published: April 23, 2025

Few-shot fine-grained image classification (FS-FGIC) intends on improving the ability to classify detailed categories with limited training samples. However, two major challenges still exist. The include effectively extracting essential features needed for while minimizing irrelevant noise, which can cause overfitting when dealing few-shot conditions. second challenge lies in achieving robust feature alignment between support and query samples, especially there are spatial variations, such as differences positions or angles of objects. This paper introduces C2-Net address these issues. innovative framework includes key modules designed overcome challenges. Cross-Layer Feature Refinement (CLFR) module has an impact quality features. It does this by blending outputs from several layers network. approach helps cut down noise at sample level. At same time, Cross-Sample Adjustment (CSFA) changes fit channel differences. makes sure that line up few Through mechanisms, reduces misalignments improves discrimination. Comprehensive experiments conducted five benchmark datasets demonstrate continously exceeds existing methods, state-of-the-art (SOTA) results most cases, improved One-shot accuracy CUB dataset 54.87% 76.51% 5-shot 79.09% 88.15%. represents a significant advancement tackling FS-FGIC.

Language: Английский

---

AI Driven Threat Detection for Autonomous Robots

Advances in computational intelligence and robotics book series, Journal Year: 2025, Volume and Issue: unknown, P. 91 - 120

Published: Feb. 21, 2025

As autonomous robots become more prevalent in diverse industries, they substantially increase productivity and safety. Still, these commonly have a passive role complex dynamic areas, which means are exposed to many possible threats. Besides security risks, cyber-attacks against damaged software, unreliable wall hardware failures, sensor issues some problems can face. These figure the creation of successful threat detection system that is only solution make sure operate safely correctly. This chapter about safety by using Artificial Intelligence (AI), part most pivotal for entire security. We explore ways AI models such as Machine Learning (ML), Deep (DL), computer vision, anomaly enable machines accurately identify react wisely

Language: Английский

---

Forecasting Wetland Transformation to Dust Source by Employing CA-Markov Model and Remote Sensing: A Case Study of Shadgan International Wetland

Wetlands, Journal Year: 2024, Volume and Issue: 44(7)

Published: Sept. 11, 2024

Language: Английский

---