Tree Crown Detection Research Articles

MangoGAN: a general adversarial network-based deep learning architecture for mango tree crown detection

We present MangoGAN, a general adversarial network (GAN)-based deep learning semantic segmentation model for the detection of mango tree crowns in remotely sensed aerial images. The aerial images are acquired by low-altitude remote sensing carried out using a quadrotor unmanned aerial vehicle in a mango orchard. Aerial images are acquired with a vision spectrum optical sensor, also popularly known as RGB images as the payload. MangoGAN is trained on 1430 images patches of size 240 × 240 pixels. The testing was carried out on 160 images. Results are analyzed using the precision, recall, F1 parameters derived from contingency matrix and by visualization using Gradcam method. The performance of the MangoGAN is compared with peer architectures trained on the same data. MangoGAN outperforms its peer architectures

How Does Sample Labeling and Distribution Affect the Accuracy and Efficiency of a Deep Learning Model for Individual Tree-Crown Detection and Delineation

Monitoring and assessing vegetation using deep learning approaches has shown promise in forestry applications. Sample labeling to represent forest complexity is the main limitation for deep learning approaches for remote sensing vegetation classification applications, and few studies have focused on the impact of sample labeling methods on model performance and model training efficiency. This study is the first-of-its-kind that uses Mask region-based convolutional neural networks (Mask R-CNN) to evaluate the influence of sample labeling methods (including sample size and sample distribution) on individual tree-crown detection and delineation. A flight was conducted over a plantation with Fokienia hodginsii as the main tree species using a Phantom4-Multispectral (P4M) to obtain UAV imagery, and a total of 2061 manually and accurately delineated tree crowns were used for training and validating (1689) and testing (372). First, the model performance of three pre-trained backbones (ResNet-34, ResNet-50, and ResNet-101) was evaluated. Second, random deleting and clumped deleting methods were used to repeatedly delete 10% from the original sample set to reduce the training and validation set, to simulate two different sample distributions (the random sample set and the clumped sample set). Both RGB image and Multi-band images derived from UAV flights were used to evaluate model performance. Each model’s average per-epoch training time was calculated to evaluate the model training efficiency. The results showed that ResNet-50 yielded a more robust network than ResNet-34 and ResNet-101 when the same parameters were used for Mask R-CNN. The sample size determined the influence of sample labeling methods on the model performance. Random sample labeling had lower requirements for sample size compared to clumped sample labeling, and unlabeled trees in random sample labeling had no impact on model training. Additionally, the model with clumped samples provides a shorter average per-epoch training time than the model with random samples. This study demonstrates that random sample labeling can greatly reduce the requirement of sample size, and it is not necessary to accurately label each sample in the image during the sample labeling process.

Comparison of Classical Methods and Mask R-CNN for Automatic Tree Detection and Mapping Using UAV Imagery

Detecting and mapping individual trees accurately and automatically from remote sensing images is of great significance for precision forest management. Many algorithms, including classical methods and deep learning techniques, have been developed and applied for tree crown detection from remote sensing images. However, few studies have evaluated the accuracy of different individual tree detection (ITD) algorithms and their data and processing requirements. This study explored the accuracy of ITD using local maxima (LM) algorithm, marker-controlled watershed segmentation (MCWS), and Mask Region-based Convolutional Neural Networks (Mask R-CNN) in a young plantation forest with different test images. Manually delineated tree crowns from UAV imagery were used for accuracy assessment of the three methods, followed by an evaluation of the data processing and application requirements for three methods to detect individual trees. Overall, Mask R-CNN can best use the information in multi-band input images for detecting individual trees. The results showed that the Mask R-CNN model with the multi-band combination produced higher accuracy than the model with a single-band image, and the RGB band combination achieved the highest accuracy for ITD (F1 score = 94.68%). Moreover, the Mask R-CNN models with multi-band images are capable of providing higher accuracies for ITD than the LM and MCWS algorithms. The LM algorithm and MCWS algorithm also achieved promising accuracies for ITD when the canopy height model (CHM) was used as the test image (F1 score = 87.86% for LM algorithm, F1 score = 85.92% for MCWS algorithm). The LM and MCWS algorithms are easy to use and lower computer computational requirements, but they are unable to identify tree species and are limited by algorithm parameters, which need to be adjusted for each classification. It is highlighted that the application of deep learning with its end-to-end-learning approach is very efficient and capable of deriving the information from multi-layer images, but an additional training set is needed for model training, robust computer resources are required, and a large number of accurate training samples are necessary. This study provides valuable information for forestry practitioners to select an optimal approach for detecting individual trees.

TCDNet: Tree Crown Detection From UAV Optical Images Using Uncertainty-Aware One-Stage Network

Tree crown detection plays a vital role in forestry management, resource statistics and yields forecasting. RGB high-resolution aerial images have emerged as a cost-effective source of data for tree crown detection. To address the challenges in the detection using UAV optical images, we propose a one-stage object detection network, TCDNet. First, the network provides an attention enhancement feature extraction module to enable the model to distinguish between tree crowns and their complex backgrounds. Second, an efficient loss is introduced to enable it to be aware of the overlap between adjacent trees, thus effectively avoiding misdetection. The experimental results on two publicly available datasets show that the proposed network outperforms state-of-art networks in terms of precision, recall and mean average precision.

Urban individual tree crown detection research using multispectral image dimensionality reduction with deep learning

Urban individual tree crown detection research using multispectral image dimensionality reduction with deep learning

A Methodology for the Automated Delineation of Crop Tree Crowns from UAV-Based Aerial Imagery by Means of Morphological Image Analysis

The popularisation of aerial remote sensing using unmanned aerial vehicles (UAV), has boosted the capacities of agronomists and researchers to offer farmers valuable data regarding the status of their crops. This paper describes a methodology for the automated detection and individual delineation of tree crowns in aerial representations of crop fields by means of image processing and analysis techniques, providing accurate information about plant population and canopy coverage in intensive-farming orchards with a row-based plant arrangement. To that end, after pre-processing initial aerial captures by means of photogrammetry and morphological image analysis, a resulting binary representation of the land plot surveyed is treated at connected component-level in order to separate overlapping tree crown projections. Then, those components are morphologically transformed into a set of seeds with which tree crowns are finally delineated, establishing the boundaries between them when they appear overlapped. This solution was tested on images from three different orchards, achieving semantic segmentations in which more than 94% of tree canopy-belonging pixels were correctly classified, and more than 98% of trees were successfully detected when assessing the methodology capacities for estimating the overall plant population. According to these results, the methodology represents a promising tool for automating the inventorying of plants and estimating individual tree-canopy coverage in intensive tree-based orchards.

Normalized Cross Correlation Template Matching for Oil Palm Tree Counting from UAV image

Tree crown detection and counting from remote sensing data such as images from Unmanned Aerial Vehicle (UAV) shows significant role in this modern era for vegetation monitoring. Since the data processing would depends on raw data available and for this case the RGB data, thus a suitable method such as template matching is presented. Normalized cross correlation is widely used as an effective measure in similarity between template image and the source or testing images. This paper focuses on six (6) steps involved in the overall process which are: (1) image acquisition, (2) template optimisation, (3) normalized cross correlation, (4) sliding window, (5) matched image and counting, and (6) accuracy assessment. Normalized cross correlation and sliding window techniques proposed for this work resulted in 80% to 89% F-measure values. This result indicates that UAV image data with appropriate image processing method/s have the potential to provide vital information for oil palm tree counting. This would be beneficial in plantation management to estimate yield and productivity. However, there are still rooms for improvement to achieve better results.

Estimating Primary Forest Attributes and Rare Community Characteristics Using Unmanned Aerial Systems (UAS): An Enrichment of Conventional Forest Inventories

The techniques for conducting forest inventories have been established over centuries of land management and conservation. In recent decades, however, compelling new tools and methodologies in remote sensing, computer vision, and data science have offered innovative pathways for enhancing the effectiveness and comprehension of these sampling designs. Now with the aid of Unmanned Aerial Systems (UAS) and advanced image processing techniques, we have never been closer to mapping forests at field-based inventory scales. Our research, conducted in New Hampshire on complex mixed-species forests, used natural color UAS imagery for estimating individual tree diameters (diameter at breast height (dbh)) as well as stand level estimates of Basal Area per Hectare (BA/ha), Quadratic Mean Diameter (QMD), Trees per Hectare (TPH), and a Stand Density Index (SDI) using digital photogrammetry. To strengthen our understanding of these forests, we also assessed the proficiency of the UAS to map the presence of large trees (i.e., >40 cm in diameter). We assessed the proficiency of UAS digital photogrammetry for identifying large trees in two ways: (1) using the UAS estimated dbh and the 40 cm size threshold and (2) using a random forest supervised classification and a combination of spectral, textural, and geometric features. Our UAS-based estimates of tree diameter reported an average error of 19.7% to 33.7%. At the stand level, BA/ha and QMD were overestimated by 42.18% and 62.09%, respectively, while TPH and SDI were underestimated by 45.58% and 3.34%. When considering only stands larger than 9 ha however, the overestimation of BA/ha at the stand level dropped to 14.629%. The overall classification of large trees, using the random forest supervised classification achieved an overall accuracy of 85%. The efficiency and effectiveness of these methods offer local land managers the opportunity to better understand their forested ecosystems. Future research into individual tree crown detection and delineation, especially for co-dominant or suppressed trees, will further support these efforts.

Benchmarking Anchor-Based and Anchor-Free State-of-the-Art Deep Learning Methods for Individual Tree Detection in RGB High-Resolution Images

Urban forests contribute to maintaining livability and increase the resilience of cities in the face of population growth and climate change. Information about the geographical distribution of individual trees is essential for the proper management of these systems. RGB high-resolution aerial images have emerged as a cheap and efficient source of data, although detecting and mapping single trees in an urban environment is a challenging task. Thus, we propose the evaluation of novel methods for single tree crown detection, as most of these methods have not been investigated in remote sensing applications. A total of 21 methods were investigated, including anchor-based (one and two-stage) and anchor-free state-of-the-art deep-learning methods. We used two orthoimages divided into 220 non-overlapping patches of 512 × 512 pixels with a ground sample distance (GSD) of 10 cm. The orthoimages were manually annotated, and 3382 single tree crowns were identified as the ground-truth. Our findings show that the anchor-free detectors achieved the best average performance with an AP50 of 0.686. We observed that the two-stage anchor-based and anchor-free methods showed better performance for this task, emphasizing the FSAF, Double Heads, CARAFE, ATSS, and FoveaBox models. RetinaNet, which is currently commonly applied in remote sensing, did not show satisfactory performance, and Faster R-CNN had lower results than the best methods but with no statistically significant difference. Our findings contribute to a better understanding of the performance of novel deep-learning methods in remote sensing applications and could be used as an indicator of the most suitable methods in such applications.

Automated tree-crown and height detection in a young forest plantation using mask region-based convolutional neural network (Mask R-CNN)

Tree-crown and height are primary tree measurements in forest inventory. Convolutional neural networks (CNNs) are a class of neural networks, which can be used in forest inventory; however, no prior studies have developed a CNN model to detect tree crown and height simultaneously. This study is the first-of-its-kind that explored training a mask region-based convolutional neural network (Mask R-CNN) for automatically and concurrently detecting discontinuous tree crown and height of Chinese fir (Cunninghamia lanceolata (Lamb) Hook) in a plantation. A DJI Phantom4-Multispectral Unmanned Aerial Vehicle (UAV) was used to obtain high-resolution images of the study site, Shunchang County, China. Tree crown and height of Chinese fir was manually delineated and derived from this UAV imagery. A portion of the ground-truthed tree height values were used as a test set, and the remaining measurements were used as the model training data. Six different band combinations and derivations of the UAV imagery were used to detect tree crown and height, respectively (Multi band-DSM, RGB-DSM, NDVI-DSM, Multi band-CHM, RGB-CHM, and NDVI-CHM combination). The Mask R-CNN model with the NDVI-CHM combination achieved superior performance. The accuracy of Chinese fir’s individual tree-crown detection was considerable (F1 score = 84.68%), the Intersection over Union (IoU) of tree crown delineation was 91.27%, and tree height estimates were highly correlated with the height from UAV imagery (R2 = 0.97, RMSE = 0.11 m, rRMSE = 4.35%) and field measurement (R2 = 0.87, RMSE = 0.24 m, rRMSE = 9.67%). Results demonstrate that the input image with an CHM achieves higher accuracy of tree crown delineation and tree height assessment compared to an image with a DSM. The accuracy and efficiency of Mask R-CNN has a great potential to assist the application of remote sensing in forests.

Individual Tree Crown Detection Using UAV Orthomosaic

Unmanned Aerial Vehicles (UAVs) are increasingly used in forestry as they are economical and flexible. This study aims to present the advantages of the drone photogrammetry method in collecting individual tree crowns, as individual tree crown detection could deliver essential ecological and economic information. The referred accuracy for individual tree crown extraction is 79.2%. Only crowns that were clearly visible were selected and manually delineated on the image because the distribution of the true crown size is significantly different from the segmented crowns. The aim of this study is to investigate UAVs orthomosaics in individual tree crown detection. The objectives of this study are to produce the orthomosaic of tree crown extraction mapping using the Pix4D software and analyze the tree crowns using tree crown delineation and the OBIA algorithm. Data processing involves the processing of aerial images using Pix4Dmapper. Automatic tree crown detection involves a tree crown delineation algorithm and OBIA operations to process the tree crown extraction. The crown delineation algorithm and OBIA algorithm operation will be compared to the actual tree crown measurement in terms of diameter and area. The tree crown delineation method obtained a 0.347m mean diameter difference from the actual tree crown diameter, while the OBIA approach obtained 4.98m. The tree crown delineation method obtained 97.26% of the actual tree crown area, while OBIA obtained 91.74%.

Individual tree crown detection from high spatial resolution imagery using a revised local maximum filtering

Accurate tree density and location are important information for optimizing the management and production of forest. Combination of remote sensing techniques and local maximum (LM) filtering algorithm provides a feasible approach to individual tree crown detection, but still faces high error under complicate canopy structure. In this study, a revised LM (RLM) algorithm is presented and evaluated for identifying individual trees from four high spatial-resolution images. Instead of a moving window technique, the RLM algorithm finds crown center seeds by searching local maximal in the transects along row and column directions of the image. Each of final crown centers is then searched using a variable window centered at the crown center seed. Strategies for splitting and merging crowns are implemented in the RLM algorithm to reduce false detection. Result showed that accuracy of the RLM algorithm was more sensitive to its minimum crown length parameter (CLmin). The RLM algorithm driven by the CLmin estimates achieved high overall accuracies between 85% and 91% and low commission (9–14%) and omission errors (8–15%) for the four images. Splitting and merging strategies implemented in the RLM algorithm effectively reduced commission and omission errors. These results indicate that the RLM algorithm is a feasible method with well-defined parameters for automatically detecting individual trees with satisfactory detection accuracy.

Individual Tree Crown Segmentation Directly from UAV-Borne LiDAR Data Using the PointNet of Deep Learning

Accurate individual tree crown (ITC) segmentation from scanned point clouds is a fundamental task in forest biomass monitoring and forest ecology management. Light detection and ranging (LiDAR) as a mainstream tool for forest survey is advancing the pattern of forest data acquisition. In this study, we performed a novel deep learning framework directly processing the forest point clouds belonging to the four forest types (i.e., the nursery base, the monastery garden, the mixed forest, and the defoliated forest) to realize the ITC segmentation. The specific steps of our approach were as follows: first, a voxelization strategy was conducted to subdivide the collected point clouds with various tree species from various forest types into many voxels. These voxels containing point clouds were taken as training samples for the PointNet deep learning framework to identify the tree crowns at the voxel scale. Second, based on the initial segmentation results, we used the height-related gradient information to accurately depict the boundaries of each tree crown. Meanwhile, the retrieved tree crown breadths of individual trees were compared with field measurements to verify the effectiveness of our approach. Among the four forest types, our results revealed the best performance for the nursery base (tree crown detection rate r = 0.90; crown breadth estimation R2 > 0.94 and root mean squared error (RMSE) < 0.2m). A sound performance was also achieved for the monastery garden and mixed forest, which had complex forest structures, complicated intersections of branches and different building types, with r = 0.85, R2 > 0.88 and RMSE < 0.6 m for the monastery garden and r = 0.80, R2 > 0.85 and RMSE < 0.8 m for the mixed forest. For the fourth forest plot type with the distribution of crown defoliation across the woodland, we achieved the performance with r = 0.82, R2 > 0.79 and RMSE < 0.7 m. Our method presents a robust framework inspired by the deep learning technology and computer graphics theory that solves the ITC segmentation problem and retrieves forest parameters under various forest conditions.

Automatic Detection and Parameter Estimation of Ginkgo biloba in Urban Environment Based on RGB Images

Individual tree crown detection and morphological parameter estimation can be used to quantify the social, ecological, and landscape value of urban trees, which play increasingly important roles in densely built cities. In this study, a novel architecture based on deep learning was developed to automatically detect tree crowns and estimate crown sizes and tree heights from a set of red‐green‐blue (RGB) images. The feasibility of the architecture was verified based on high‐resolution unmanned aerial vehicle (UAV) images using a neural network called FPN‐Faster R‐CNN, which is a unified network combining a feature pyramid network (FPN) and a faster region‐based convolutional neural network (Faster R‐CNN). Among more than 400 tree crowns, including 213 crowns of Ginkgo biloba, in 7 complex test scenes, 174 ginkgo tree crowns were correctly identified, yielding a recall level of 0.82. The precision and F‐score were 0.96 and 0.88, respectively. The mean absolute error (MAE) and mean absolute percentage error (MAPE) of crown width estimation were 0.37 m and 8.71%, respectively. The MAE and MAPE of tree height estimation were 0.68 m and 7.33%, respectively. The results showed that the architecture is practical and can be applied to many complex urban scenes to meet the needs of urban green space inventory management.

Individual Tree-Crown Detection and Species Classification in Very High-Resolution Remote Sensing Imagery Using a Deep Learning Ensemble Model

Traditional methods for individual tree-crown (ITC) detection (image classification, segmentation, template matching, etc.) applied to very high-resolution remote sensing imagery have been shown to struggle in disparate landscape types or image resolutions due to scale problems and information complexity. Deep learning promised to overcome these shortcomings due to its superior performance and versatility, proven with reported detection rates of ~90%. However, such models still find their limits in transferability across study areas, because of different tree conditions (e.g., isolated trees vs. compact forests) and/or resolutions of the input data. This study introduces a highly replicable deep learning ensemble design for ITC detection and species classification based on the established single shot detector (SSD) model. The ensemble model design is based on varying the input data for the SSD models, coupled with a voting strategy for the output predictions. Very high-resolution unmanned aerial vehicles (UAV), aerial remote sensing imagery and elevation data are used in different combinations to test the performance of the ensemble models in three study sites with highly contrasting spatial patterns. The results show that ensemble models perform better than any single SSD model, regardless of the local tree conditions or image resolution. The detection performance and the accuracy rates improved by 3–18% with only as few as two participant single models, regardless of the study site. However, when more than two models were included, the performance of the ensemble models only improved slightly and even dropped.

Individual Tree Detection from UAV Imagery Using Hölder Exponent

This article explores the application of Hölder exponent analysis for the identification and delineation of single tree crowns from very high-resolution (VHR) imagery captured by unmanned aerial vehicles (UAV). Most of the present individual tree crown detection (ITD) methods are based on canopy height models (CHM) and are very effective as far as an accurate digital terrain model (DTM) is available. This prerequisite is hard to accomplish in some environments, such as alpine forests, because of the high tree density and the irregular topography. Indeed, in such conditions, the photogrammetrically derived DTM can be inaccurate. A novel image processing method supports the segmentation of crowns based only on the parameter related to the multifractality description of the image. In particular, the multifractality is related to the deviation from a strict self-similarity and can be treated as the information about the level of inhomogeneity of considered data. The multifractals, even if well established in image processing and recognized by the scientific community, represent a relatively new application in VHR aerial imagery. In this work, the Hölder exponent (one of the parameters related to multifractal description) is applied to the study of a coniferous forest in the Western Alps. The infrared dataset with 10 cm pixels is captured by a UAV-mounted optical sensor. Then, the tree crowns are detected by a basic workflow. This consists of the thresholding of the image on the basis of the Hölder exponent. Then, the single crowns are segmented through a multiresolution segmentation approach. The ITD segmentation was validated through a two-level validation analysis that included a visual evaluation and the computing of quantitative measures based on 200 reference crowns. The results were checked against the ITD performed in the same area but using only spectral, textural, and elevation information. Specifically, the visual assessment included the estimation of the producer’s and user’s accuracies and the F1 score. The quantitative measures considered are the root mean square error (RMSE) (for the area, the perimeter, and the distance between centroids) and the over-segmentation and under-segmentation indices, the Jaccard index, and the completeness index. The F1 score indicates positive results (over 73%) as well as the completeness index that does not exceed 0.23 on a scale of 0 to 1, taking 0 as the best result possible. The RMSE of the extension of crowns is 3 m2, which represents only 14% of the average extension of reference crowns. The performance of the segmentation based on the Hölder exponent outclasses those based on spectral, textural, and elevation information. Despite the good results of the segmentation, the method tends to under-segment rather than over-segment, especially in areas with sloping. This study lays the groundwork for future research into ITD from VHR optical imagery using multifractals.

Individual tree crown detection and delineation across a woodland using leaf-on and leaf-off imagery from a UAV consumer-grade camera

Extraction of information about individual trees using remotely sensed data is essential to supporting ecological and commercial applications in forest environments. Data acquired by consumer-grade cameras onboard unmanned aerial vehicles (UAV) offer an affordable option of high-spatial resolution imagery that can be used to extract forest structural information at a tree level. The aim of this work is to investigate the potential and accuracy of UAV time-series data to automatically detect and delineate tree crowns across an entire woodland. The workflow (presented in a step-by-step manner) involves the construction of a canopy height model (CHM) based on digital elevation models derived from the UAV photogrammetric point clouds. A watershed-based approach is modified to automatically detect and delineate the tree crowns, based on the CHM and the brightness information from the UAV orthomosaics. The accuracy of the proposed method was evaluated by comparing its results against manually delineated tree crowns. The results show an overall accuracy of 63%, where conifer species were more accurately delineated (up to 80%), while broadleaf species returned lower accuracies (<50 % ). Continued research is necessary to improve the confidence of automated individual tree crown detection and delineation, especially over complex forest structures.

Street tree segmentation from mobile laser scanning data

ABSTRACT Mobile laser scanning (MLS) measurement technology is well-suited for individual-tree-based urban forest inventory because it can capture a dense three-dimensional (3D) point cloud from vertical structures of roadside trees in a rapid and cost-effective way. Street tree segmentation from MLS data is a preliminary and essential step for further extraction and analysis of tree attributes and is also a challenging problem because of the complex street environment. The accuracy of current methods is limited by information loss introduced by voxelization and artificially designed detection and segmentation rules based on prior knowledge that more or less lack ability to distinguish between street trees and other urban objects. To address the above issues, a two-step strategy for street tree segmentation method from MLS data was proposed, which worked on points directly instead of a set of voxels generated by spatial partition and made use of tree crown and trunk detectors trained by supervised learning algorithm. First, 16 local statistical features, including width, depth, elevation, dimensionality, and density features, were extracted from the sphere domain of each point using the grid index. Then these features were fused and tree crown and trunk detectors were trained from labelled MLS data of street through Discrete AdaBoost algorithm. Second, based on frame projection of point-wise prediction results of tree crown and trunk detectors, individual trees were located and segmented by exploiting non-connectivity between adjacent trees and adjacency within single trees. Experimental results have demonstrated that the proposed method can obtain point-wise street tree segmentation from MLS data with high accuracy tree crown and trunk detection without tedious detection or segmentation rules. The error rates of crown detector on test set were less than 0.3% accompanied by over 99.33% recall rates and over 98.59% precision rates and the error rates of trunk detector on test set were less than 0.06% accompanied by over 92.44% recall rates and over 94.14% precision rates.

Tree Crown Delineation Algorithm Based on a Convolutional Neural Network

Tropical forests concentrate the largest diversity of species on the planet and play a key role in maintaining environmental processes. Due to the importance of those forests, there is growing interest in mapping their components and getting information at an individual tree level to conduct reliable satellite-based forest inventory for biomass and species distribution qualification. Individual tree crown information could be manually gathered from high resolution satellite images; however, to achieve this task at large-scale, an algorithm to identify and delineate each tree crown individually, with high accuracy, is a prerequisite. In this study, we propose the application of a convolutional neural network—Mask R-CNN algorithm—to perform the tree crown detection and delineation. The algorithm uses very high-resolution satellite images from tropical forests. The results obtained are promising—the R e c a l l , P r e c i s i o n , and F 1 score values obtained were were 0.81 , 0.91 , and 0.86 , respectively. In the study site, the total of tree crowns delineated was 59,062 . These results suggest that this algorithm can be used to assist the planning and conduction of forest inventories. As the algorithm is based on a Deep Learning approach, it can be systematically trained and used for other regions.

Airborne Tree Crown Detection for Predicting Spatial Heterogeneity of Canopy Transpiration in a Tropical Rainforest

Tropical rainforests comprise complex 3D structures and encompass heterogeneous site conditions; their transpiration contributes to climate regulation. The objectives of our study were to test the relationship between tree water use and crown metrics and to predict spatial variability of canopy transpiration across sites. In a lowland rainforest of Sumatra, we measured tree water use with sap flux techniques and simultaneously assessed crown metrics with drone-based photogrammetry. We observed a close linear relationship between individual tree water use and crown surface area (R2 = 0.76, n = 42 trees). Uncertainties in predicting stand-level canopy transpiration were much lower using tree crown metrics than the more conventionally used stem diameter. 3D canopy segmentation analyses in combination with the tree crown–water use relationship predict substantial spatial heterogeneity in canopy transpiration. Among our eight study plots, there was a more than two-fold difference, with lower transpiration at riparian than at upland sites. In conclusion, we regard drone-based canopy segmentation and crown metrics to be very useful tools for the scaling of transpiration from tree- to stand-level. Our results indicate substantial spatial variation in crown packing and thus canopy transpiration of tropical rainforests.