Comparison of Tree Typologies Mapping Using Random Forest Classifier Algorithm of PRISMA and Sentinel-2 Products in Different Areas of Central Italy

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(3), P. 356 - 356

Published: Jan. 22, 2025

The continuous development of satellite imagery, coupled with advancements in machine learning technologies, allows detailed mapping terrestrial landscapes. This study evaluates the classification performance tree typologies using Sentinel-2 and PRISMA data, focusing on central Italy’s different areas. purpose is to assess role spectral spatial resolution land cover classification, contributing forest management conservation efforts. Random Forest Classifier was applied classify across two areas: Roman Coastal region Lake Vico Basin. Ground truth (GT) collected from a trial citizen survey campaign, were used for training validation. datasets, particularly when processed PCA, consistently outperformed Sentinel-2. PCA dataset achieved highest overall accuracy 71.09% 87.15% Basin, emphasizing value resolution. However, showed comparative strength spatially heterogeneous Tree more uniform distribution, such as hazelnut chestnut, higher compared mixed-species forests. assesses that remains viable alternative where critical also considering limited images’ availability. Moreover, work explores potential combining satellites accurate GT improved mapping.

Language: Английский

---

Innovative Decision Fusion for Accurate Crop/Vegetation Classification with Multiple Classifiers and Multisource Remote Sensing Data

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(9), P. 1579 - 1579

Published: April 29, 2024

Obtaining accurate and real-time spatial distribution information regarding crops is critical for enabling effective smart agricultural management. In this study, innovative decision fusion strategies, including Enhanced Overall Accuracy Index (E-OAI) voting the Index-based Majority Voting (OAI-MV), were introduced to optimize use of diverse remote sensing data various classifiers, thereby improving accuracy crop/vegetation identification. These strategies utilized integrate classification outcomes from distinct feature sets (including Gaofen-6 reflectance, Sentinel-2 time series vegetation indices, biophysical variables, Sentinel-1 backscatter coefficients, their combinations) using classifiers (Random Forests (RFs), Support Vector Machines (SVMs), Maximum Likelihood (ML), U-Net), taking two grain-producing areas (Site #1 Site #2) in Haixi Prefecture, Qinghai Province, China, as research area. The results indicate that employing U-Net on feature-combined yielded highest overall (OA) 81.23% 91.49% #2, respectively, single classifier experiments. E-OAI strategy, compared original OAI boosted OA by 0.17% 6.28%. Furthermore, OAI-MV strategy achieved 86.02% 95.67% respective study sites. This highlights strengths features discerning different crop types. Additionally, proposed effectively harness benefits multisource features, significantly enhancing classification.

Language: Английский

---

Integration of Hyperspectral Imaging and AI Techniques for Crop Type Mapping: Present Status, Trends, and Challenges

Remote Sensing, Journal Year: 2025, Volume and Issue: 17(9), P. 1574 - 1574

Published: April 29, 2025

Accurate and efficient crop maps are essential for decision-makers to improve agricultural monitoring management, thereby ensuring food security. The integration of advanced artificial intelligence (AI) models with hyperspectral remote sensing data, which provide richer spectral information than multispectral imaging, has proven highly effective in the precise discrimination types. This systematic review examines evolution platforms, from Unmanned Aerial Vehicle (UAV)-mounted sensors space-borne satellites (e.g., EnMAP, PRISMA), explores recent scientific advances AI methodologies mapping. A protocol was applied identify 47 studies databases peer-reviewed publications, focusing on sensors, input features, classification architectures. analysis highlights significant contributions Deep Learning (DL) models, particularly Vision Transformers (ViTs) hybrid architectures, improving accuracy. However, also identifies critical gaps, including under-utilization limited multi-sensor need modeling approaches such as Graph Neural Networks (GNNs)-based methods geospatial foundation (GFMs) large-scale type Furthermore, findings highlight importance developing scalable, interpretable, transparent maximize potential imaging (HSI), underrepresented regions Africa, where research remains limited. provides valuable insights guide future researchers adopting HSI reliable mapping, contributing sustainable agriculture global

Language: Английский

---

In-Depth Analysis and Characterization of a Hazelnut Agro-Industrial Context through the Integration of Multi-Source Satellite Data: A Case Study in the Province of Viterbo, Italy

Remote Sensing, Journal Year: 2024, Volume and Issue: 16(7), P. 1227 - 1227

Published: March 30, 2024

The production of “Nocciola Romana” hazelnuts in the province Viterbo, Italy, has evolved into a highly efficient and profitable agro-industrial system. Our approach is based on hierarchical framework utilizing aggregated data from multiple temporal sources, offering valuable insights spatial, temporal, phenological distributions hazelnut crops To achieve our goal, we harnessed power Google Earth Engine utilized collections satellite images Sentinel-2 Sentinel-1. By creating dense stack multi-temporal images, precisely mapped groves area. During testing phase model pipeline, achieved an F1-score 99% by employing Hierarchical Random Forest algorithm conducting intensive sampling using high-resolution imagery. Additionally, employed clustering process to further characterize identified areas. Through this process, unveiled distinct regions exhibiting diverse spectral, responses. We successfully delineated actual extent cultivation, totaling 22,780 hectares, close accordance with national statistics, which reported 23,900 hectares total 21,700 for year 2022. In particular, three geographic distribution patterns orchards confined within PDO (Protected Designation Origin)-designated region. methodology pursued, years aggregate one SAR spectral separation approach, effectively allowed identification specific perennial crop, enabling deeper characterization various aspects influenced environmental configurations agronomic practices.The accurate mapping open opportunities implementing precision agriculture strategies, thereby promoting sustainability maximizing yields thriving

Language: Английский

---

Identification of High-Photosynthetic-Efficiency Wheat Varieties Based on Multi-Source Remote Sensing from UAVs

Agronomy, Journal Year: 2024, Volume and Issue: 14(10), P. 2389 - 2389

Published: Oct. 16, 2024

Breeding high-photosynthetic-efficiency wheat varieties is a crucial link in safeguarding national food security. Traditional identification methods necessitate laborious on-site observation and measurement, consuming time effort. Leveraging unmanned aerial vehicle (UAV) remote sensing technology to forecast photosynthetic indices opens up the potential for swiftly discerning varieties. The objective of this research develop multi-stage predictive model encompassing nine indicators at field scale breeding. These include soil plant analyzer development (SPAD), leaf area index (LAI), net rate (Pn), transpiration (Tr), intercellular CO2 concentration (Ci), stomatal conductance (Gsw), photochemical quantum efficiency (PhiPS2), PSII reaction center excitation energy capture (Fv’/Fm’), quenching coefficient (qP). ultimate goal differentiate through model-based predictions. This gathered red, green, blue spectrum (RGB) multispectral (MS) images eleven stages jointing, heading, flowering, filling. Vegetation (VIs) texture features (TFs) were extracted as input variables. Three machine learning regression models (Support Vector Machine Regression (SVR), Random Forest (RF), BP Neural Network (BPNN)) employed construct across multiple growth stages. Furthermore, conducted principal component analysis (PCA) membership function on predicted values optimal each indicator, established comprehensive evaluation high efficiency, cluster screen test materials. categorized into three groups, with SH06144 Yannong 188 demonstrating higher efficiency. moderately efficient group comprises Liangxing 19, SH05604, SH06085, Chaomai 777, SH05292, Jimai 22, Guigu 820, totaling seven Xinmai 916 Jinong 114 fall category lower aligning closely results clustering based actual measurements. findings suggest that employing UAV-based multi-source identify feasible. study provide theoretical basis winter phenotypic monitoring breeding using sensing, offering valuable insights advancement smart practices

Language: Английский

---