Article
Environmental Sciences
Zhihui Mao, Zhuo Lu, Yanjie Wu, Lei Deng
Summary: This study explores the capability of 2D image-based texture and spectrum features in estimating the diameter at breast height (DBH) of individual trees. The results show that there is a strong correlation between texture features and DBH, and using texture features alone provides the highest accuracy in DBH estimation.
Article
Environmental Sciences
Kaisen Ma, Zhenxiong Chen, Liyong Fu, Wanli Tian, Fugen Jiang, Jing Yi, Zhi Du, Hua Sun
Summary: This study compares the methods based on canopy height model (CHM) and normalized point cloud (NPC) for individual tree segmentation in UAV-LiDAR data, as well as tree height parameter extraction from nine plots of three forest types. The results show that the NPC-based methods outperformed the CHM-based methods in individual tree segmentation, and the type and complexity of a forest affect the accuracy of the segmentation.
Article
Forestry
Martin Slavik, Karel Kuzelka, Roman Modlinger, Peter Surovy
Summary: This study proposes a method of tree species classification using individual tree metrics derived from three-dimensional point cloud data obtained by unmanned aerial vehicle laser scanning. The metrics of 1045 trees were evaluated using a generalized linear model and random forest techniques, leading to automated assignment of individual trees into either coniferous or broadleaf groups. The inclusion of a spatial aggregation index called the Clark-Evans index significantly improved classification accuracy, with overall accuracies of 94.8% and 95.1% achieved using the generalized linear model and random forest approaches, respectively.
Article
Environmental Sciences
Jeremy Arkin, Nicholas C. Coops, Lori D. Daniels, Andrew Plowright
Summary: Accurate estimation of forest canopy fuels is crucial for wildfire prediction and mitigation. This study examines the ability of LiDAR point clouds from RPAS to characterize the vertical arrangement and volume of crown fuels, showing good match between extracted and manually measured clusters but a tendency for overprediction of lower boundaries in the automated method.
Article
Environmental Sciences
Aaron M. Sparks, Mark Corrao, Alistair M. S. Smith
Summary: This study evaluated the accuracy of seven individual tree detection methods in coniferous forest stands, showing that higher ALS pulse density data resulted in higher ITD accuracy. Omission errors were mainly related to stand density, and the use of simple canopy height model methods could reduce omission errors.
Article
Environmental Sciences
Qingda Chen, Tian Gao, Jiaojun Zhu, Fayun Wu, Xiufen Li, Deliang Lu, Fengyuan Yu
Summary: Accurate individual tree segmentation is important for forest management and the study of forest ecosystems. This study introduced a new UAV-LiDAR dataset, FULD, which fused leaf-off and leaf-on point clouds, to assess its benefits for tree segmentation and height estimation in dense deciduous forests. The results showed that the combination of FULD and the layer stacking segmentation algorithm produced the highest accuracies across all forest types and improved tree height estimation.
Article
Forestry
Aaron M. Sparks, Alistair M. S. Smith
Summary: In this study, the ability of the ForestView(R) algorithm to detect individual trees, classify tree species, live/dead status, canopy position, and estimate height and DBH in a complex forest was assessed. The algorithm showed high accuracy in stands with lower canopy cover but lower accuracy in stands with higher canopy cover.
Article
Engineering, Multidisciplinary
Can Vatandaslar, Mustafa Zeybek
Summary: Using handheld mobile laser scanning (HMLS) for forest inventory (FI) surveys, with the assistance of machine learning and innovative algorithms, efficiently estimates and maps key FI parameters. HMLS-derived data show strong correlation with field reference data at the single-tree level.
Article
Forestry
Pawel Hawrylo, Jaroslaw Socha, Piotr Wezyk, Wojciech Ochal, Wojciech Krawczyk, Jakub Miszczyszyn, Luiza Tyminska-Czabanska
Summary: This paper presents a universal method for determining forest top height (TH) based on ALS data and evaluates its accuracy through an experiment. The results show that the individual tree detection approach is the most accurate and can meet the needs of forest practitioners and researchers.
FOREST ECOLOGY AND MANAGEMENT
(2024)
Article
Environmental Sciences
Julia Tatum, David Wallin
Summary: This study is the first to model tree species from LiDAR in natural Pacific Northwest forests and to classify these species at the landscape scale. The results suggest that LiDAR alone can provide useful information on tree species in limited applications, even in structurally challenging conditions. With slight changes to the modeling approach, higher accuracies may be possible.
Article
Environmental Sciences
Melissa Latella, Fabio Sola, Carlo Camporeale
Summary: This study introduces a novel LiDAR algorithm for accurate individual tree detection in deciduous forests, with low sensitivity to parameter setup and applicability in low-density point cloud analysis. Additionally, the algorithm demonstrates potential for more complex tools in forest modeling and management.
Article
Forestry
Katrina Ariel Henn, Alicia Peduzzi
Summary: The benefits and services of urban forests, especially carbon storage, are well documented. A generalizable individual urban tree model was developed using NAIP aerial imagery and LiDAR data. The model was then used to estimate the total biomass and carbon stored for all the trees in the county. Recommendations include adapting ground inventory techniques to the limitations of remote sensing biomass estimation.
Article
Environmental Sciences
Zhenyang Hui, Shuanggen Jin, Dajun Li, Yao Yevenyo Ziggah, Bo Liu
Summary: This paper proposed an individual tree extraction method based on transfer learning and Gaussian mixture model separation, which achieved an average correctness of 87.68% in experiments, outperforming two classical methods.
Article
Ecology
Louise Terryn, Kim Calders, Markku Akerblom, Harm Bartholomeus, Mathias Disney, Shaun Levick, Niall Origo, Pasi Raumonen, Hans Verbeeck
Summary: Detailed 3D quantification of tree structure is crucial for understanding tree- and plot-level biophysical processes. Our ITSMe toolbox, which works with LiDAR tree point clouds and quantitative structure models, provides a robust framework for obtaining individual tree structural metrics from 3D data. It is open-source and aims to make the use of 3D data more straightforward and transparent for researchers interested in tree structure information.
METHODS IN ECOLOGY AND EVOLUTION
(2023)
Article
Forestry
Yao Liu, Haotian You, Xu Tang, Qixu You, Yuanwei Huang, Jianjun Chen
Summary: Individual structural parameters of trees, such as height and biomass, are crucial for monitoring forest resources. This study investigated the crown segmentation accuracy of different tree species using three-dimensional (3D) data obtained from LiDAR and image-derived points. The results revealed that LiDAR data generally outperformed image-derived 3D data in crown segmentation accuracy. Among the tested segmentation algorithms, PointNet++ achieved the highest accuracy, while LSS yielded the lowest. The tree species also influenced crown segmentation, with Liriodendron chinense showing the best segmentation and Ficus microcarpa showing the worst. These findings emphasize the importance of considering 3D data source, segmentation algorithm, and tree species when conducting individual tree crown segmentation.
Article
Environmental Sciences
Bo Chen, Fang Qiu, Bingfang Wu, Hongyue Du
Article
Geography, Physical
Yuhong Zhou, Fang Qiu, Feng Ni, Yifei Lou, Caiyun Zhang, Mohammed Alfarhan, Ali A. Al-Dosari
GISCIENCE & REMOTE SENSING
(2016)
Article
Remote Sensing
Yuhong Zhou, Fang Qiu, Ali A. Al-Dosari, Mohammed S. Alfarhan
INTERNATIONAL JOURNAL OF REMOTE SENSING
(2015)
Article
Remote Sensing
Caiyun Zhang, Hannah Cooper, Donna Selch, Xuelian Meng, Fang Qiu, Soe W. Myint, Charles Roberts, Zhixiao Xie
REMOTE SENSING LETTERS
(2014)
Article
Environmental Sciences
Yunwei Tang, Linhai Jing, Fan Shi, Xiao Li, Fang Qiu
Article
Environmental Sciences
Hu Ding, Kai Liu, Xiaozheng Chen, Liyang Xiong, Guoan Tang, Fang Qiu, Josef Strobl
Article
Environmental Sciences
Fan Shi, Fang Qiu, Xiao Li, Yunwei Tang, Ruofei Zhong, Cankun Yang
Article
Environmental Sciences
Haoming Wan, Yunwei Tang, Linhai Jing, Hui Li, Fang Qiu, Wenjin Wu
Summary: The spatial distribution of forest stands plays a crucial role in understanding and managing forests. The fusion of multiple remote sensing data sources, including high-spatial-resolution images, time-series images, and LiDAR data, is essential for accurately identifying tree species for forest stand classification. The FSP algorithm, based on curve matching, has been developed to effectively fuse and analyze these data sources, outperforming traditional machine learning classification methods in terms of accuracy and stability.
Article
Geography, Physical
Yunwei Tang, Fang Qiu, Bangjin Wang, Di Wu, Linhai Jing, Zhongchang Sun
Summary: Recent developments in deep learning have introduced new methods for land cover classification. However, most methods neglect the spatial association of land cover classes in remote sensing images. This research proposes a deep relearning method based on recurrent neural networks, which improves classification accuracy by considering the spatial arrangement of land cover classes.
GISCIENCE & REMOTE SENSING
(2022)
Article
Geography, Physical
Yunwei Tang, Fang Qiu, Linhai Jing, Fan Shi, Xiao Li
ISPRS JOURNAL OF PHOTOGRAMMETRY AND REMOTE SENSING
(2020)
Article
Environmental Sciences
Fan Shi, Fang Qiu, Xiao Li, Ruofei Zhong, Cankun Yang, Yunwei Tang
Article
Criminology & Penology
Bryan Chastain, Fang Qiu, Alex R. Piquero
AMERICAN JOURNAL OF CRIMINAL JUSTICE
(2016)
Article
Geography
Fang Qiu, Bryan Chastain, Yuhong Zhou, Caiyun Zhang, Harini Sridharan
Article
Geography
Harini Sridharan, Fang Qiu
GEOGRAPHICAL ANALYSIS
(2013)
