Article
Biodiversity Conservation
Franziska Wolff, Tiina H. M. Kolari, Miguel Villoslada, Teemu Tahvanainen, Pasi Korpelainen, Pedro A. P. Zamboni, Timo Kumpula
Summary: Plant communities of mires are important for ecological processes such as carbon storage and gas fluxes. Mapping mire vegetation using UAVs can provide valuable information for ecosystem assessment. However, accurate mapping of plant communities remains challenging due to overlapping spectral signatures of plant species.
ECOLOGICAL INDICATORS
(2023)
Article
Environmental Sciences
Youming Zhang, Na Ta, Song Guo, Qian Chen, Longcai Zhao, Fenling Li, Qingrui Chang
Summary: The use of a fast and accurate unmanned aerial vehicle (UAV) digital camera platform to estimate the leaf area index (LAI) of kiwifruit orchards is significant for growth monitoring, yield estimation, and field management. This study used high-resolution UAV images to extract spectral and textural parameters, which were then used to construct regression models for LAI estimation. The results showed that the model combining texture features had better prediction accuracy compared to the model based solely on spectral indices.
Article
Forestry
Chao Gao, Honglei Lin, Haiqing Hu
Summary: Forest-fire-risk was predicted in the Heihe area of Heilongjiang Province, China using random forest (RF) and backpropagation neural network (BPNN) algorithms. The driving factors correlated with forest-fire occurrence were analyzed and 11 factors were found to have a significant correlation. The prediction accuracy and goodness of fit of RF and BPNN algorithms were similar, indicating that both methods are suitable for forest-fire occurrence prediction. High-fire-risk zones were mainly located in the northwestern and central parts of the Heihe area.
Article
Environmental Sciences
Zhangyu Sun, Bao Zhang, Yibin Yao
Summary: This study uses machine learning methods to improve the estimation of empirical T-m in China, significantly increasing accuracy and reducing variations in accuracy across space and time.
Article
Engineering, Environmental
Yuelei Xu, Yan Huang, Zhongyang Guo
Summary: This study analyzed the impact of meteorological elements, AOD products, and modeling methods on the accuracy of the AOD-PM2.5 model, finding that incorporating meteorological elements that vary with time and height can significantly improve model accuracy in eastern China, Terra AOD products have a higher product index compared to Aqua AOD products, and the RF model outperforms other modeling methods in terms of performance.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2021)
Article
Agriculture, Multidisciplinary
Luis Padua, Alessando Matese, Salvatore Filippo Di Gennaro, Raul Morais, Emanuel Peres, Joaquim J. Sousa
Summary: Vineyard classification is crucial for decision-support systems, and data acquired by unmanned aerial vehicles and sensors can improve classification performance. In this study, different machine learning methods were employed to classify vineyard elements, and the results showed that using multiple features and the artificial neural network (ANN) approach achieved the best classification performance.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Energy & Fuels
Hector Felipe Mateo Romero, Luis Hernandez-Callejo, Miguel angel Gonzalez Rebollo, Valentin Cardenoso-Payo, Victor Alonso Gomez, Jose Ignacio Morales Aragones, Ranganai Tawanda Moyo
Summary: This work presents a method to predict the output power of Photovoltaic cells using Electroluminescence images. It compares different machine learning methods, from traditional ones like Random Forest and Gradient Boosting to deep learning methods like Recurrent Neural Networks and Convolutional Neural Networks. The paper also addresses the issue of unbalanced data and attempts to solve it using Synthetic Images created by a Generative Adversarial Network. The results show that the Gradient-Boosting based method with a pre-trained ResNet50 as a feature extraction method performs the best with a Mean Absolute Error of 0.0341 and a Mean Squared Error of 0.00211.
Article
Automation & Control Systems
Yiming Lei, Haiping Zhu, Junping Zhang, Hongming Shan
Summary: This study proposes a method for medical image classification with ordinal labels, which improves the model's generalization ability by combining convolutional neural networks and differential forests in a meta-learning framework. The key components of the method are the tree-wise weighting network and the grouped feature selection module. Experimental results demonstrate the superior performance of this method over existing state-of-the-art methods on two medical image classification datasets with ordinal labels.
IEEE-CAA JOURNAL OF AUTOMATICA SINICA
(2022)
Article
Biodiversity Conservation
Huifang Zhang, Zhonggang Tang, Binyao Wang, Baoping Meng, Yu Qin, Yi Sun, Yanyan Lv, Jianguo Zhang, Shuhua Yi
Summary: This study aims to provide a non-destructive method for quickly obtaining data on grassland aboveground biomass (AGB) using unmanned aerial vehicles (UAVs) and machine learning techniques. By comparing different index combinations, it was found that the model combining horizontal and vertical indices performed the best. However, the lack of vegetation height information in areas with high vegetation coverage remains a limitation.
GLOBAL ECOLOGY AND CONSERVATION
(2022)
Article
Agriculture, Multidisciplinary
Xinbing Wang, Yuxin Miao, Rui Dong, Krzysztof Kusnierek
Summary: This study aimed to determine the potential to minimize the differences of two commonly used active canopy sensors (ACSs) in maize nitrogen (N) status diagnosis and recommendation using multi-source data fusion and machine learning. The results showed that the use of multi-source data fusion with machine learning models can improve the accuracy of ACS-based N status diagnosis and recommendation and minimize the performance differences of different active sensors.
PRECISION AGRICULTURE
(2023)
Article
Engineering, Electrical & Electronic
Qiang Fu
Summary: This paper proposes a multimodal feature fusion supervised RGB-D image saliency detection network, which uses a dual-stream side-supervision module and a multimodal feature fusion module to improve the quality and robustness of saliency prediction results.
JOURNAL OF SENSORS
(2022)
Article
Chemistry, Analytical
Jianhang Zhang, Shucheng Huang, Jingting Li, Yan Wang, Zizhao Dong, Su-Jing Wang
Summary: This paper proposes a portable wireless transmission system for multi-channel acquisition of surface electromyography (EMG) signals. By placing electrodes around the face, this system can detect muscle activity in 16 regions simultaneously. It employs wireless transmission technology to increase portability. Experimental results showed that this system is reliable and practical for recognizing facial movements.
Article
Geosciences, Multidisciplinary
Moritz Lange, Henri Suominen, Mona Kurppa, Leena Jarvi, Emilia Oikarinen, Rafael Savvides, Kai Puolamaki
Summary: Using regression models to replicate air pollutant concentrations in urban boulevards, this study found that log-linear regression performs best and most robustly on new independent data. The study also demonstrated the importance of detecting concept drift and avoiding overfitting when selecting models and features.
GEOSCIENTIFIC MODEL DEVELOPMENT
(2021)
Article
Environmental Sciences
Haikuan Feng, Huilin Tao, Zhenhai Li, Guijun Yang, Chunjiang Zhao
Summary: This study explores the use of unmanned aerial vehicles equipped with RGB and hyperspectral cameras for monitoring crop growth. It combines multiple growth indicators to estimate a comprehensive growth index (CGI) and finds that spectral indices are more strongly correlated with the CGI than single growth-monitoring indicators. The multiple linear regression (MLR) method produces the best CGI estimates. Using hyperspectral indices provides more accurate CGI estimations compared to using RGB-image indices.
Article
Multidisciplinary Sciences
W. K. V. J. B. Kulasooriya, R. S. S. Ranasinghe, Udara Sachinthana Perera, P. Thisovithan, I. U. Ekanayake, D. P. P. Meddage
Summary: This study investigates the importance of utilizing explainable artificial intelligence (XAI) on various machine learning (ML) models to predict the strength characteristics of basalt-fiber reinforced concrete (BFRC). It introduces three tree-based ML models and two explanation methods to provide interpretable explanations for the predictions. The findings highlight the need for further XAI-based research in concrete strength prediction and involvement of domain experts to evaluate XAI results.
SCIENTIFIC REPORTS
(2023)