Ecological Informatics最新文献

A comparison of deep neural network compression for citizen-driven tick and mosquito surveillance 基于深度神经网络压缩的蜱蚊监测比较

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-08 DOI: 10.1016/j.ecoinf.2025.103437

Yichao Liu , Emmanuel Dufourq , Peter Fransson , Joacim Rocklöv

引用次数: 0

Soil moisture content of northern peatlands from close-range spectral data 来自近距离光谱数据的北方泥炭地土壤水分含量

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-08 DOI: 10.1016/j.ecoinf.2025.103466

Susanna Karlqvist, Jussi Juola, Aarne Hovi, Sini-Selina Salko, Iuliia Burdun, Miina Rautiainen

{"title":"Soil moisture content of northern peatlands from close-range spectral data","authors":"Susanna Karlqvist, Jussi Juola, Aarne Hovi, Sini-Selina Salko, Iuliia Burdun, Miina Rautiainen","doi":"10.1016/j.ecoinf.2025.103466","DOIUrl":"10.1016/j.ecoinf.2025.103466","url":null,"abstract":"<div><div>Peatlands play a critical role in the global carbon cycle. Their carbon exchange functions are highly sensitive to moisture conditions and water table levels, which are increasingly threatened by climate change and land-use modifications. While satellite remote sensing enables large-scale monitoring of peatland moisture conditions, precise close-range measurements are essential for calibrating and validating these methods. Previous close-range studies have typically focused on data from a limited number of peatland sites, creating gaps in the understanding of soil surface moisture estimation across diverse peatland types and climatic zones. Our study addressed these gaps using close-range hyperspectral field measurements from 13 northern peatlands spanning hemiboreal to Arctic regions in Estonia and Finland. We evaluated multiple techniques, including spectral indices, continuum removal, full-spectrum analysis, and Continuous Wavelet Transform (CWT), with Kernel Partial Least Squares (KPLS) regression. Additionally, we compared hyperspectral methods with simulated multispectral data. Our results showed that hyperspectral data with CWT processing provided the most accurate soil moisture estimations across diverse peatland environments (R<sup>2</sup> = 0.65 and RMSE = 17.7 %). Additionally, the model based on multispectral bands, achieved moderate prediction accuracy (R<sup>2</sup> = 0.53 and RMSE = 20.4 %), which was competitive with full-spectrum analysis (R<sup>2</sup> = 0.58 and RMSE = 19.4 %). Separating peatlands into minerotrophic and ombrotrophic categories further improved prediction accuracy, particularly for the minerotrophic sites (with CWT: R<sup>2</sup> = 0.74 and RMSE = 15.9 %). In contrast, spectral indices performed poorly (R<sup>2</sup> ≤ 0.26 and RMSE ≥ 25.8 %), suggesting that they may be unsuitable for large-scale remote sensing applications.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103466"},"PeriodicalIF":7.3,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270106","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Retrieval of planetary boundary layer height from CALIPSO satellite observations using a machine learning approach 利用机器学习方法从CALIPSO卫星观测资料中检索行星边界层高度

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-08 DOI: 10.1016/j.ecoinf.2025.103431

A. Salcedo-Bosch , F. Rocadenbosch , C.I. Argañaraz , G. Curci , S. Lolli

{"title":"Retrieval of planetary boundary layer height from CALIPSO satellite observations using a machine learning approach","authors":"A. Salcedo-Bosch , F. Rocadenbosch , C.I. Argañaraz , G. Curci , S. Lolli","doi":"10.1016/j.ecoinf.2025.103431","DOIUrl":"10.1016/j.ecoinf.2025.103431","url":null,"abstract":"<div><div>The planetary boundary layer height (PBLH) is a key variable in air quality, climate modeling, and weather prediction. Traditional retrieval methods, such as radiosondes, provide high accuracy but lack spatial coverage. This study presents a Random Forest (RF) model based on Machine Learning (ML) to estimate PBLH from ten years of Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) on board the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO), using radiosonde measurements as a reference. The model achieves an <span><math><msup><mrow><mi>R</mi></mrow><mrow><mn>2</mn></mrow></msup></math></span> of 0.67 and an RMSE of 278.02 m with a spatial resolution of <span><math><mo>≈</mo></math></span> 20 × 20 km<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> in a test set that covers mainly Europe and North America. Unlike previous methods, our approach does not require atmospheric typing and uses minimal data filtering, demonstrating robustness under diverse aerosol and cloud conditions. Although validation is currently limited to mid-latitude regions, the method offers a scalable approach to global monitoring and supports the management of climate and air quality. Future work will extend the validation to other geographic zones and explore deep learning models for further improvements.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103431"},"PeriodicalIF":7.3,"publicationDate":"2025-10-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270103","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Recent advances in dinoflagellate cyst: Integrating review with visual taxonomic perspectives 鞭毛囊肿的最新研究进展：结合视觉分类的观点

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-07 DOI: 10.1016/j.ecoinf.2025.103460

Wenjing Guo , Qisheng Yu , Yue Yu , Youyi Tang , Nan Zeng , Chia-Ying Anderin Chuang , Yuelu Jiang

{"title":"Recent advances in dinoflagellate cyst: Integrating review with visual taxonomic perspectives","authors":"Wenjing Guo , Qisheng Yu , Yue Yu , Youyi Tang , Nan Zeng , Chia-Ying Anderin Chuang , Yuelu Jiang","doi":"10.1016/j.ecoinf.2025.103460","DOIUrl":"10.1016/j.ecoinf.2025.103460","url":null,"abstract":"<div><div>Resting cysts are a critical stage in the life cycle of certain dinoflagellates, involving active growth, sexual reproduction, and dormancy. With the global rise in harmful algal blooms (HABs), interest in the ecological roles of these cysts has grown. This review systematically analyzed literatures indexed in the Web of Science Core Collection and identified three major themes within the field of modern dinoflagellate cyst research: field investigations, life history and ecological implications, and taxonomy and phylogenetics. Advances in these areas have improved our understanding of how resting cysts contribute to reconstructing eutrophication histories, assessing invasion risks, exploring genetic diversity, and supporting HAB forecasting. This review also outlines recent progress in dinoflagellate classification, emphasizing developments in both morphological and molecular approaches, and highlights emerging topics related to toxin-producing species and genetic markers. Together, these insights provide an integrated view of how cyst formation, dormancy, and germination influence ecological adaptation, population dynamics, bloom management, and evolutionary processes, while identifying directions for future research.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103460"},"PeriodicalIF":7.3,"publicationDate":"2025-10-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270101","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Predicting wildfire occurrences in Portugal using machine learning classification models 使用机器学习分类模型预测葡萄牙的野火发生情况

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-06 DOI: 10.1016/j.ecoinf.2025.103455

Jorge Caiado , Mariana Marques

{"title":"Predicting wildfire occurrences in Portugal using machine learning classification models","authors":"Jorge Caiado , Mariana Marques","doi":"10.1016/j.ecoinf.2025.103455","DOIUrl":"10.1016/j.ecoinf.2025.103455","url":null,"abstract":"<div><div>Wildfires pose significant environmental and socio-economic challenges, particularly in fire-prone regions such as Portugal. The ability to predict wildfire occurrences is essential for improving preparedness and mitigation strategies. This study evaluates the effectiveness of three machine learning classification models (Logistic Regression, Random Forest and XGBoost) in forecasting wildfire occurrences across four Portuguese districts: Lisbon, Porto, Setúbal and Viseu. Using historical fire occurrence data and meteorological variables, the models were trained and tested on different land-use categories, including settlements, brush and agriculture. The results indicate that brush fires are the most predictable due to strong climatic influences, with models achieving F1-scores above 0.93. Settlement fires, in contrast, were more challenging to predict, likely due to human-driven variability, whereas agricultural fires exhibited intermediate predictability. To address dataset imbalances, the Synthetic Minority Oversampling Technique (SMOTE) was applied, leading to improvements in recall but a trade-off in precision. Feature importance analysis highlights the influence of long-term temporal trends, meteorological conditions and human activity on wildfire risk. These findings demonstrate the potential of machine learning models in wildfire forecasting and provide valuable insights for policymakers and fire management authorities in designing targeted prevention strategies.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103455"},"PeriodicalIF":7.3,"publicationDate":"2025-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270102","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Estimating seasonal fractional green and dead vegetation cover in rehabilitated ecosystems using drone remote sensing 利用无人机遥感估算恢复生态系统的季节性绿植被和死植被覆盖度

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-06 DOI: 10.1016/j.ecoinf.2025.103456

Matthew Noonan, Yuhong He, Daniel Nelson, Haoxuan Ge, Ryan Siu, Tim P. Duval

{"title":"Estimating seasonal fractional green and dead vegetation cover in rehabilitated ecosystems using drone remote sensing","authors":"Matthew Noonan, Yuhong He, Daniel Nelson, Haoxuan Ge, Ryan Siu, Tim P. Duval","doi":"10.1016/j.ecoinf.2025.103456","DOIUrl":"10.1016/j.ecoinf.2025.103456","url":null,"abstract":"<div><div>Restoration of ecosystems affected by surface mining requires effective monitoring to evaluate rehabilitation success. Conventional approaches, including field surveys and satellite remote sensing, are often constrained by cost, spatial resolution, and temporal frequency, particularly when monitoring multiple small or fragmented sites. To address these limitations, this study uses Unmanned Aerial Vehicles (UAV) acquired visible imagery to estimate within-season variation in fractional green and dead vegetation cover across three sites at different rehabilitation stages in Southern Ontario, Canada. Using monthly UAV imagery collected between May and August 2023, and random forest regression models, we generated high-resolution maps of fractional green and dead vegetation cover for each site. Green vegetation cover was estimated with high accuracy (R<sup>2</sup> = 0.94–0.97; RMSE = 8–10 %) across the three sites, while dead vegetation cover was predicted with moderate to high accuracy (R<sup>2</sup> = 0.74–0.95; RMSE = 10–12 %). Mapping results revealed that site two (rehabilitated in 2006) showed limited recovery, with low green vegetation cover (24–47 % monthly average) and high dead vegetation cover (52–76 % monthly average), likely due to poor substrate and minimal follow-up. Site one (rehabilitated in 2017) demonstrated strong seasonal greening despite similar rehabilitated treatment (48–60 % monthly average) and moderate dead vegetation cover (40–51 % monthly average). Site three (rehabilitated in 2022) maintained high green cover (48–56 % monthly average) and low dead materials (10–15 % monthly average), suggesting early recovery, possibly supported by favorable conditions or better restoration inputs. These findings demonstrate the utility of UAV-derived visible imagery for cost-effective, fine-scale monitoring of vegetation dynamics in rehabilitated ecosystems.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103456"},"PeriodicalIF":7.3,"publicationDate":"2025-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270104","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A novel methodology for Explainable Artificial Intelligence integrated with geostatistics for air pollution control and environmental management 可解释人工智能的新方法与空气污染控制和环境管理的地质统计学相结合

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-03 DOI: 10.1016/j.ecoinf.2025.103450

Mateusz Zareba, Tomasz Danek

{"title":"A novel methodology for Explainable Artificial Intelligence integrated with geostatistics for air pollution control and environmental management","authors":"Mateusz Zareba, Tomasz Danek","doi":"10.1016/j.ecoinf.2025.103450","DOIUrl":"10.1016/j.ecoinf.2025.103450","url":null,"abstract":"<div><div>This study focuses on understanding air pollution dynamics through the integration of Explainable Artificial Intelligence (XAI) and spatio-temporal geostatistics for air pollution control. It combines the classic approach with the simultaneous evaluation of multiple machine learning (ML) models to enhance the interpretability of model decisions. Data from a dense network of 52 IoT (internet of things) sensors deployed across an urban municipality and surrounding areas were analyzed to identify key predictors of air pollution trends. XAI is employed to interpret complex relationships between pollution levels, meteorological conditions, human activity patterns, and geographic factors. Spatio-temporal geostatistics serve as the foundation for interpreting XAI in both spatial and temporal contexts. July exhibited the least spatial variability in PM<sub>2.5</sub> concentrations, with the hour of the day being the key predictor (13.04% of predictive importance), while December showed the highest spatial variability, with atmospheric pressure as the dominant factor (13.84% of predictive importance). Precipitation was the least influential predictor in both months (3.00 %). Four and nine distinct clusters with significant spatial variability in predictor importance were identified. Analysis of transition matrices revealed both stable and dynamic clusters, highlighting the complex nature of PM<sub>2.5</sub> emissions as they vary between warm and cold seasons, characteristic of moderate climate zones. This methodology enhances the understanding of air pollution dynamics and provides a transparent framework for urban pollution control and management decision-making. The findings contribute to the development of smart urban management strategies, supporting sustainable city planning and pollution mitigation efforts, while advocating for the right of citizens to a cleaner environment.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103450"},"PeriodicalIF":7.3,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Mapping forage nutritional quality in temperate natural grasslands using interpretable machine learning 利用可解释机器学习绘制温带自然草原牧草营养质量图

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-03 DOI: 10.1016/j.ecoinf.2025.103452

Fusen Nan , Xingxin Lu , Tianci Chen , Jianfei Yu , Xiaoqian Yang , Yanming Gong , Kaihui Li , Xiao-Bo Wang

{"title":"Mapping forage nutritional quality in temperate natural grasslands using interpretable machine learning","authors":"Fusen Nan , Xingxin Lu , Tianci Chen , Jianfei Yu , Xiaoqian Yang , Yanming Gong , Kaihui Li , Xiao-Bo Wang","doi":"10.1016/j.ecoinf.2025.103452","DOIUrl":"10.1016/j.ecoinf.2025.103452","url":null,"abstract":"<div><div>Accurate spatial prediction of forage nutritional quality is critical for optimizing livestock production and supporting regional food security planning. This study developed an interpretable machine learning framework based on Random Forest (RF) and Support Vector Machine (SVM) algorithms to generate high-resolution maps of key forage nutritional parameters—crude protein (CP), ether extract (EE), acid detergent fiber (ADF), and neutral detergent fiber (NDF)—across natural grasslands in the Ili River Valley, Xinjiang, China. Our approach integrated 40 environmental covariates and explicitly accounted for both immediate and legacy climate effects through time-lag and accumulation analyses. Predictor selection was optimized using forward feature selection combined with <em>k</em>-fold nearest-neighbor distance matching cross-validation to reduce overfitting. Results indicated that climate variables significantly influenced forage quality via time-lag and accumulation effects, with distinct drivers identified for nutrient content versus pools: CP and EE content were primarily influenced by climate, whereas ADF and NDF content as well as all nutrient pools were predominantly driven by vegetation traits. Model performance varied considerably, with nutrient pools (R<sup>2</sup> = 0.63–0.73) being predicted more accurately than nutrient content (R<sup>2</sup> = 0.42–0.58). Spatially, high-quality forage was concentrated in mountain meadows and temperate meadow steppes, while low-quality forage was more prevalent in temperate desert steppes. Uncertainty analysis revealed higher prediction errors in desert steppes, likely due to limited sampling density and extreme environmental conditions. These findings offer a scientific basis for targeted grassland management and underscore the importance of appropriate model selection and covariate inclusion for reliable spatial predictions of forage quality.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103452"},"PeriodicalIF":7.3,"publicationDate":"2025-10-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145223195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Testing social network metrics as proxies for governance performance: A simulation-based experiment in watershed management 测试社会网络指标作为治理绩效的代理：流域管理中基于模拟的实验

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-10-01 DOI: 10.1016/j.ecoinf.2025.103442

Nelson Jatel

{"title":"Testing social network metrics as proxies for governance performance: A simulation-based experiment in watershed management","authors":"Nelson Jatel","doi":"10.1016/j.ecoinf.2025.103442","DOIUrl":"10.1016/j.ecoinf.2025.103442","url":null,"abstract":"<div><div>This study introduces a simulation-based modelling framework to systematically evaluate whether widely used social network analysis (SNA) metrics function as credible proxies for governance performance. I generated 100 synthetic governance networks with covariance structures linking collaboration, equity, resilience, participation, and coordination to structural properties. A suite of analyses, including multiple regression models, permutation tests, partial correlations, and hierarchical clustering, was applied to test the predictive validity of reciprocity, transitivity, Gini degree, k-core, betweenness centrality, clustering coefficient, modularity, and density. Results demonstrate reproducible structural–functional linkages: reciprocity and transitivity robustly predict collaboration, equity is inversely tied to Gini degree, and resilience depends on k-core prominence and betweenness centrality. The modelling workflow, implemented in Python with open scripts and datasets, provides transparent benchmarks for interpreting governance-relevant network metrics. Beyond advancing theory, this framework enhances the diagnostic utility of SNA, supporting more reliable decision-support tools for watershed governance and environmental management. By embedding governance processes into a reproducible, simulation-based workflow, this study extends the reach of ecological informatics beyond biophysical systems to include social structures that shape environmental outcomes. The approach provides transferable benchmarks and open-source resources that strengthen reproducibility, comparability, and integration of governance diagnostics within ecological informatics research.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103442"},"PeriodicalIF":7.3,"publicationDate":"2025-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Wavelet convolution and multi-scale attention for Chinese Herbal Medicine recognition 小波卷积与多尺度关注在中药识别中的应用

IF 7.3 2区环境科学与生态学

Ecological Informatics Pub Date : 2025-09-29 DOI: 10.1016/j.ecoinf.2025.103449

Ruchan Dong , Bin Jiang , Wenjing Wu , Zhengrui Zhao , Hongjie Zhao , Qianyan Shen , Ya Liu , Haobin Luo

{"title":"Wavelet convolution and multi-scale attention for Chinese Herbal Medicine recognition","authors":"Ruchan Dong , Bin Jiang , Wenjing Wu , Zhengrui Zhao , Hongjie Zhao , Qianyan Shen , Ya Liu , Haobin Luo","doi":"10.1016/j.ecoinf.2025.103449","DOIUrl":"10.1016/j.ecoinf.2025.103449","url":null,"abstract":"<div><div>The intelligent and precise recognition of Chinese Herbal Medicine is a critical challenge for advancing smart healthcare, particularly as artificial intelligence integrates with the modernization of traditional Chinese medicine. Existing methods struggle to model the multi-scale morphological characteristics of Chinese Herbal Medicine and achieve high classification accuracy under complex backgrounds. To address these limitations, we constructed the Chinese Herbal Medicine Multi-Morphology Image Dataset-50 (CHM-Morph-50), comprising images of 50 species with annotations for leaf texture, three-dimensional contour, and graded background complexity. Leveraging this dataset, we propose the Wavelet Convolution and Multi-Scale Attention Network (WCMSA-Net), which uses wavelet convolution to decompose images into high- and low-frequency components for parallel extraction of complementary details, combined with a multi-scale spatial pyramid attention mechanism to dynamically capture discriminative features. An improved label-smoothing focal loss with class-weight adjustment is introduced to mitigate class imbalance. Experimental results show that WCMSA-Net achieves a Top-1 accuracy of 88.6% on CHM-Morph-50, exceeding ResNet50 and MobileNet by 3.4% and 10.4%, respectively, and maintaining robustness under occlusion and blur. This study offers a high-accuracy, scalable framework for the digital recognition of Chinese Herbal Medicine, with potential applications in the identification of other medicinal plant species.</div></div>","PeriodicalId":51024,"journal":{"name":"Ecological Informatics","volume":"92 ","pages":"Article 103449"},"PeriodicalIF":7.3,"publicationDate":"2025-09-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145270292","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0