Article
Chemistry, Multidisciplinary
Ling Dai, Guangyun Zhang, Jinqi Gong, Rongting Zhang
Summary: This paper proposes a data-driven method for hyperspectral remotely sensed data, which can autonomously extract key features and interactively learn feature indexes, providing a more flexible and creative framework compared to traditional methods.
APPLIED SCIENCES-BASEL
(2021)
Article
Environmental Sciences
Zhen Shu, Xiangyun Hu, Hengming Dai
Summary: This paper proposes a click-based interactive building extraction method in remote sensing images, utilizing boundary information and progress guidance map, combined with convolutional neural network to achieve accurate extraction. Experiments demonstrate that the proposed method outperforms existing methods in terms of stability and accuracy.
Article
Remote Sensing
Hongzhang Xu, Hongjie He, Ying Zhang, Lingfei Ma, Jonathan Li
Summary: This study compares the performance differences of 12 commonly used loss functions in road segmentation tasks in remote sensing imagery. It is found that the region-based loss function type generally performs better than the distribution-based one in terms of F1, IoU, and road segmentation maps, while the compound loss function type is comparable to the region-based one. This paper aims to provide suggestions for choosing the loss function that best suits the purposes of road segmentation-related studies.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2023)
Article
Geochemistry & Geophysics
Zhenghua Huang, Zifan Zhu, Qing An, Zhicheng Wang, Qin Zhou, Tianxu Zhang, Ali Saleh Alshomrani
Summary: This letter proposes a novel enhancement framework for remotely sensed images to correct luminance guided by weighted least squares (WLS), which separates the image into base and detail layers for enhancement. Experimental results show that the proposed method outperforms current techniques in contrast improvement and detail preservation.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Environmental Sciences
Xiangkai Wang, Yong Xue, Chunlin Jin, Yuxin Sun, Na Li
Summary: Accurate calculation of near surface ozone concentration with high spatial resolution data is crucial for addressing severe ozone pollution and health impact assessment. This study improved the downscaling algorithm based on mutual information and successfully achieved the downsampling of TROPOMI ground O-3 concentration data from 30 km to 1 km in China. By combining with the surface O-3 concentration data obtained by the LightGBM algorithm and AOD data from MODIS, the spatial resolution of ozone concentration in the ground layer has been significantly increased, resulting in improved precision.
FRONTIERS IN ENVIRONMENTAL SCIENCE
(2022)
Article
Engineering, Electrical & Electronic
Zhenghua Huang, Lei Wang, Qing An, Qin Zhou, Hanyu Hong
Summary: This paper proposes a new enhancement framework for remotely sensed images called Global-Local Enhancement Network (GLE-Net). The framework corrects the intensity of the images by learning extra information from collected training data, improving both the low-frequency and detail components, and producing high-quality images. The GLE-Net method performs well in preserving brightness and fine details, outperforming existing techniques.
IEEE SIGNAL PROCESSING LETTERS
(2022)
Article
Geosciences, Multidisciplinary
Zekun Gao, Yutong Jiang, Junyu He, Jiaping Wu, George Christakos
Summary: In this study, multiple types of ocean temperature data were fused using the Bayesian maximum entropy method, and four fusion products were generated and compared with other datasets. The results demonstrated the high accuracy and potential of the BME method in marine data interpolation and fusion.
SPATIAL STATISTICS
(2023)
Article
Computer Science, Interdisciplinary Applications
Aleksandar Anzel, Dominik Heider, Georges Hattab
Summary: This study proposes juxtaposed Taylor and Mutual Information Diagrams, which combine statistics, information theory, and data visualization, to help track and summarize the performance of complex models. The library provides interactive implementation of these diagrams and supports both continuous and categorical attributes.
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Remote Sensing
Xiaoyun Xiang, Dongsheng Xiao
Summary: A method utilizing remote sensing indicators to assess the socioeconomic development of poverty-stricken counties was proposed and found to be accurate in evaluating the development level of these counties. The study showed an overall increase in ESDPC values of selected impoverished counties between 2013 and 2017. Validation with a comprehensive development index model confirmed the effectiveness and reliability of this method, demonstrating the potential for using remotely sensed indicators in assessing regional socioeconomic development.
EUROPEAN JOURNAL OF REMOTE SENSING
(2021)
Article
Engineering, Civil
Mohit Kumar, Reet Kamal Tiwari, Kamal Kumar, Kuldeep Singh Rautela
Summary: In this study, a statistical analysis of MODIS snow time series data and the Seasonal Autoregressive Integrated Moving Average (SARIMA) model were applied to accurately measure cyclic snow accumulation and depletion in the Beas river basin in the Himalayan region from 2003 to 2018. The Box-Jenkins methodology was used to forecast snow accumulation and depletion based on seasonality, stationarity, ACF, and PACF plots, as well as maximum likelihood estimation and diagnostic checking. The forecasting models for snow accumulation period (October-February) and snow depletion period (March-September) showed good agreement with observed data, with R-2 values of 0.83 and 0.89, respectively. This research highlights the potential of using satellite data and statistical modeling for monitoring snow cover in remote and inaccessible regions.
AQUA-WATER INFRASTRUCTURE ECOSYSTEMS AND SOCIETY
(2023)
Article
Environmental Studies
Pietro De Marinis, Samuele De Petris, Filippo Sarvia, Giacinto Manfron, Evelyn Joan Momo, Tommaso Orusa, Gianmarco Corvino, Guido Sali, Enrico Mondino Borgogno
Summary: The study focuses on existing agricultural production systems in the eastern Democratic Republic of Congo, using remote sensing imagery and entropy analysis to investigate the spatial distribution of subsistence-oriented agriculture (SOA) and business-oriented agriculture (BOA) in the Katoyi collectivity of Masisi territory. The results show that land use and entropy analysis can provide updated information on existing land distribution among different production systems, supporting better intervention strategies in development cooperation and pro-poor agrarian land tenure reforms in conflict-ridden landscapes.
Proceedings Paper
Computer Science, Artificial Intelligence
Zhaohui Wang
Summary: In an Improved K-means clustering algorithm, initial clustering time can be saved by making initial division based on previous clustering results and maintaining the relationship among stable classes. Clustering lossless compression algorithm can efficiently eliminate the interspectral and intra-spectral redundancy at high convergent speed through enhancing intra-class redundancy. The comparison of the parameter analysis of the AVIRIS images with other lossless compression algorithms shows that this clustering lossless compression algorithm is more efficient.
2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA)
(2021)
Article
Agriculture, Multidisciplinary
Gustavo Togeiro de Alckmin, Lammert Kooistra, Richard Rawnsley, Arko Lucieer
Summary: This study compared the performance of canopy-based technique and spectral vegetation indices in pasture biomass estimation, finding that the canopy-based technique outperformed spectral vegetation indices while the selected vegetation indices in combination with different regression techniques improved accuracy and precision.
PRECISION AGRICULTURE
(2021)
Article
Geography, Physical
D. Jaskierniak, A. Lucieer, G. Kuczera, D. Turner, P. N. J. Lane, R. G. Benyon, S. Haydon
Summary: Estimation of forest stocking density per hectare is crucial for understanding forest dynamics post disturbances, and in this study, a novel bottom-up approach for individual tree and crown delineation (ITCD) using UAS LiDAR technology was developed and evaluated across 39 flight sites. The algorithm achieved a high mean F-score of 0.91 and accurately estimated the forest stocking density across the sites.
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2021)
Article
Remote Sensing
Samuel Hillman, Luke Wallace, Arko Lucieer, Karin Reinke, Darren Turner, Simon Jones
Summary: In recent years, Unoccupied Aircraft Systems (UAS) have been utilized for capturing detailed forest structure information with high resolution and accuracy. The data collected from UAS platforms, especially UAS LiDAR point clouds, contain valuable information describing fuel properties in different vegetation layers, providing important insights for forest fuel assessment and fire hazard evaluation.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2021)
Article
Environmental Sciences
Irfan A. Iqbal, Jon Osborn, Christine Stone, Arko Lucieer
Summary: Digital aerial photogrammetry (DAP) is considered as a cost-effective alternative to airborne laser scanning (ALS) for forest inventory, showing similar accuracy. Individual tree detection algorithms have been developed from ALS or DAP data, but the application of ITDs to DAP-based point clouds has not been reported. Results show agreement between ALS- and DAP-based ITD results, with the number of trees per hectare having the greatest influence on tree detection rates.
Review
Environmental Sciences
Juan C. Montes-Herrera, Emiliano Cimoli, Vonda Cummings, Nicole Hill, Arko Lucieer, Vanessa Lucieer
Summary: Monitoring marine ecosystems requires observations of attributes at different scales which traditional methods struggle to provide. Proximal optical sensing methods bridge this observational gap by tracking changes non-invasively. Underwater hyperspectral imaging shows potential for monitoring pigmentation and identifying minerals at small spatial scales.
Article
Forestry
B. K. Yadav, A. Lucieer, G. J. Jordan, S. C. Baker
Summary: This study tested the predictive power of landscape topography and geology on vegetation density in a wet eucalypt forest, finding that geological and topographic attributes can provide useful predictions for vegetation layers with 30 m DTM resolution. The model's predictive accuracy can potentially be further tested on a larger geographical area using lower-density LiDAR point clouds and medium-resolution satellite data.
AUSTRALIAN FORESTRY
(2022)
Article
Agriculture, Multidisciplinary
Gustavo Togeirode Alckmin, Arko Lucieer, Richard Rawnsley, Lammert Kooistra
Summary: Frequent biomass measurement is important for optimal perennial ryegrass management in dairy operations. Development of accurate and automated technological solutions for biomass assessment is vital. UAVs with multispectral cameras can help deploy machine learning algorithms for real-time biomass mapping, but radiometric calibration and generalization of models need improvement.
COMPUTERS AND ELECTRONICS IN AGRICULTURE
(2022)
Article
Ecology
Peter A. Harrison, Nicolo Camarretta, Sean Krisanski, Tanya G. Bailey, Neil J. Davidson, Glen Bain, Rowena Hamer, Riana Gardiner, Kirstin Proft, Mohammad Sadegh Taskhiri, Paul Turner, Darren Turner, Arko Lucieer
Summary: Using remote sensing technologies can assist in ecological restoration of forests at various levels, from observing structural complexity and animal behavior at the community level, monitoring vegetation structure and ecosystem services at plot level, to accurately classifying plants and showing genetic variations at the individual level. However, challenges remain to be addressed to promote wider use of remote sensing in restoration efforts.
ECOLOGICAL MANAGEMENT & RESTORATION
(2021)
Article
Environmental Sciences
Poornima Sivanandam, Arko Lucieer
Summary: Effective methods for tree delineation and species classification in an Australian native forest were identified in this study. The study found that the highest classification accuracies were achieved at the superpixel scale, and the DeepForest method showed potential for tree detection compared to conventional methods.
Article
Environmental Sciences
Leonard Hambrecht, Arko Lucieer, Zbynek Malenovsky, Bethany Melville, Ana Patricia Ruiz-Beltran, Stuart Phinn
Summary: Remotely sensing morphological traits can assess functional diversity of forests regardless of spatial scale. Trait probability density (TPD) is a computationally intensive method for calculating functional diversity, but using kernel density estimator (KDE) is more efficient than one-class support vector machine (SVM) when the number of input traits is high. Dimension reduction techniques and appropriate kernel size are recommended for optimizing TPD calculations.
Article
Ecology
David M. J. S. Bowman, Stefania Ondei, Arko Lucieer, Scott Foyster, Lynda D. Prior
Summary: The study investigates the boundaries between forests and sedgelands in western Tasmania and finds that they have been geographically stable over historical timeframes. Keystone resprouter species contribute to the rapid recovery of vegetation after fire.
Article
Environmental Sciences
Ben Weeding, Arko Lucieer, Peter T. Love, Tom Remenyi, Rebecca M. B. Harris
Summary: As the Earth's climate warms, the frequency and severity of outdoor thermal conditions that threaten human life are increasing. To effectively prepare and adapt to this challenge, it is crucial to understand both the current baseline thermal conditions and how they will change in the future. However, current efforts to measure and model baseline thermal conditions have not fully considered the contributions of radiation and have been conducted at coarse temporal and spatial resolutions.
Review
Geography
Dipendra Bhattarai, Arko Lucieer, Heather Lovell, Jagannath Aryal
Summary: Night-time light (NTL) satellite imagery provides unique insights into the energy sector. However, there is a limited number of studies reviewing the relationship between electricity consumption and NTL. This paper aims to systematically review these studies and finds a large variability in regression performance, indicating a need for further refinement in remote sensing techniques and approaches.
Article
Ecology
Darren Turner, Emiliano Cimoli, Arko Lucieer, Ryan S. Haynes, Krystal Randall, Melinda J. Waterman, Vanessa Lucieer, Sharon A. Robinson
Summary: This study develops a model to predict water content in Antarctic moss beds using laboratory experiments and spectroscopy analysis. The model is then applied to high-resolution images taken by unmanned aerial systems (UAS) to monitor water content in different conditions. The study demonstrates the potential of UAS-borne short-wave infrared (SWIR) imaging for mapping and quantifying water content in Antarctic moss beds.
REMOTE SENSING IN ECOLOGY AND CONSERVATION
(2023)
Article
Environmental Sciences
Francois du Toit, Nicholas C. Coops, Blaise Ratcliffe, Yousry A. El-Kassaby, Arko Lucieer
Summary: Coastal Douglas-fir is an economically important softwood species in North America. The use of remote sensing technology allows for the measurement of branching traits and the estimation of attributes such as branch length, angle, width, and volume. The study found that branch angle had the highest heritability, while tree height and branch length had the highest genetic correlation, indicating the importance of considering branch-level metrics in breeding programs.
SCIENCE OF REMOTE SENSING
(2023)