Tree Counts Research Articles

Implementing Deep Learning algorithms for urban tree detection and geolocation with high-resolution aerial, satellite, and ground-level images

Urban forests are becoming increasingly important for human well-being as they provide ecosystem services that contribute to improving well-being of city dwellers and to addressing climate change. However, despite their importance, there is an information gap in most of the world's urban forests due to the high cost and complexity of conducting standard forest inventories in urban environments. New technologies based on artificial intelligence can represent a smart and efficient alternative to costly traditional inventories. In this paper, we present an approach based on deep learning algorithms for the detection, counting, and geopositioning of trees using a combination of ground-level and aerial/satellite imagery. We tested several convolutional networks, exploring different combinations of hyperparameters and adjusting the query distance between ground-level images, detection radius, and various resolutions of satellite and aerial images. Our methodology is able to detect and accurately locate 79% of the urban street tree with a positional accuracy of 60 cm to the center of the canopy. Additionally, this approach allows us to determine the availability of photographs of urban trees, indicating from which Google Street View image each tree is visible. Our research provides a scalable and replicable solution to the scarcity of urban tree data and information worldwide, demonstrating the potential of artificial intelligence to revolutionize the way in which we inventory and monitor urban forests.

Strong above-ground impacts of a non-native ungulate do not cascade to impact below-ground functioning in a boreal ecosystem.

1. Experimental studies across biomes demonstrate that herbivores can have significant effects on ecosystem functioning. Herbivore effects, however, can be highly variable with studies demonstrating positive, neutral or negative relationships between herbivore presence and different components of ecosystems. Mixed effects are especially likely in the soil, where herbivore effects are largely indirect mediated through effects on plants. 2. We conducted a long-term experiment to disentangle the effects of non-native moose in boreal forests on plant communities, nutrient cycling, soil composition and soil organism communities. 3. To explore the effect of moose on soils, we conduct separate analyses on the soil organic and mineral horizons. Our data come from 11 paired exclosure-control plots in eastern and central Newfoundland, Canada that provide insight into 22-25 years of moose herbivory. We fit piecewise structural equations models (SEM) to data for the organic and mineral soil horizons to test different pathways linking moose to above-ground and below-ground functioning. 4. The SEMs revealed that moose exclusion had direct positive impacts on adult tree count and an indirect negative impact on shrub percent cover mediated by adult tree count. We detected no significant impact of moose on soil microbial C:N ratio or net nitrogen mineralization in the organic or mineral soil horizon. Soil temperature and moisture, however, was more than twice as variable in the presence (i.e. control) than absence (i.e. exclosure) of moose. Overall, we observed clear impacts of moose on above-ground forest components with limited indirect effects below-ground. Even after 22-25 years of exclusion, we did not find any evidence of moose impacts on soil microbial C:N ratio and net nitrogen mineralization. 5. Our long-term study and mechanistic path analysis demonstrates that soils can be resilient to ungulate herbivore effects despite evidence of strong effects above-ground. Long-term studies and analyses such as this one are relatively rare yet critical for reconciling some of the context-dependency observed across studies of ungulates effects on ecosystem functions. Such studies may be particularly valuable in ecosystems with short growing seasons such as the boreal forest.

Soft Segmentation of Terrestrial Laser Scanning Point Cloud of Forests

As the three-dimensional (3D) laser scanner is widely used for forest inventory, analyzing and processing point cloud data captured with a 3D laser scanner have become an important research topic in recent years. The extraction of single trees from point cloud data is essential for further investigation at the individual tree level, such as counting trees and trunk analysis, and many developments related to this topic have been published. However, constructing an accurate and automated method to obtain the tree crown silhouette from the point cloud data is challenging because the tree crowns often overlap between adjacent trees. A soft segmentation method that uses K-Nearest Neighbor (KNN) and contour shape constraints at the overlap region is proposed to solve this task. Experimental results show that the visual effect of the tree crown shape and the precision of point cloud segmentation have improved. It is concluded that the proposed method works well for tree crown segmentation and silhouette reconstruction from the terrestrial laser scanning point cloud data of the forest.

AI-Accelerated Decision Making in Forest Management

This paper is a continuation of the research on the application of artificial intelligence in counting trees with the use of methods for the automatic analysis of photogrammetric data in forests of the continental region. This paper is devoted to the AI application in accelerating decision making processes in forest management. It also discusses how RGB imagery from drones could replace aerial and satellite hyperspectral imagery and automatically detect unhealthy and dead trees. Experimental research was conducted to verify whether Faster R-CNN can automatically detect and classify snag and trees weakened by diseases on aerial RGB data, enabling a quick response to forest-threatening factors. The research is based on photogrammetric data taken in areas of forest districts subordinate to the Regional Directorate of State Forests in Zielona Góra. Non-metric imagery data was collected from drones and small airplanes with a photogrammetric container and postprocessed with respect to the photogrammetric constraints. The results show that in specific cases aerial and satellite hyperspectral imagery can be replaced by RGB orthomosaics in order to decrease the time needed for forestry treatments.

Ranking top-k trees in tree-based phylogenetic networks.

Tree-based phylogenetic networks provide a powerful model for representing complex data or non-tree-like evolution. Such networks consist of an underlying evolutionary tree called a "support tree" (also known as a "subdivision tree") together with extra arcs added between the edges of that tree. However, a tree-based network can have exponentially many support trees, and this leads to a variety of computational problems. Recently, Hayamizu established a theory called the structure theorem for rooted binary phylogenetic networks and provided linear-time and linear-delay algorithms for different problems, such as counting, optimization, and enumeration of support trees. However, in practice, it is often more useful to search for both optimal and near-optimal solutions than to calculate only an optimal solution. In the present paper, we thus consider the following problem: Given a tree-based phylogenetic network N where each arc is weighted by its probability, compute the ranking of top-k support trees of N according to their likelihood values. We provide a linear-delay (and hence optimal) algorithm for this problem.

Environment factors influencing abundance and growth of wild yam in broadleaved forest, Tsirang District, Bhutan

Yams (Dioscorea) consist of approximately 600 species in the world. Yams are considered a staple food in Asia, tropical and subtropical regions. Genetic erosion and unsustainable harvesting practices are among the main problems associated with wild yam plants. This study was conducted with the objectives to enumerate wild yam species composition and environmental parameters affecting its growth and abundance. Systematic sampling of the study site was carried out based on different land aspects. Line transects were laid out in selected land aspects within the elevation range of 825 – 1935 m. Five plots in each transect with a 100 m plot interval comprising seven transects were laid out in the study area. Each plot was divided into sub-plot of 20 x 20 m, 5 x 5 m, and 1 x 1 m to enumerate trees, shrubs, and herbs respectively. Seven species of yams were found from the broadleaved forest and spearman correlation showed negative association of yam abundance with tree counts and canopy closure. The areas with higher density of trees had low count of yam abundance. Yam grows well in open canopy in lower altitudes comparing to high altitudes. Yam abundance was positively related with shrubs, herbs, soil moisture, soil organic matter, soil organic carbon, and nitrogen content. Analysis of Variance (ANOVA) showed significant difference of yam abundance in 7 different transects due to different ecological parameters influencing the abundance in different transects though transect being in a broadleaved forest.

Deep learning enables image-based tree counting, crown segmentation, and height prediction at national scale.

Sustainable tree resource management is the key to mitigating climate warming, fostering a green economy, and protecting valuable habitats. Detailed knowledge about tree resources is a prerequisite for such management but is conventionally based on plot-scale data, which often neglects trees outside forests. Here, we present a deep learning-based framework that provides location, crown area, and height for individual overstory trees from aerial images at country scale. We apply the framework on data covering Denmark and show that large trees (stem diameter >10 cm) can be identified with a low bias (12.5%) and that trees outside forests contribute to 30% of the total tree cover, which is typically unrecognized in national inventories. The bias is high (46.6%) when our results are evaluated against all trees taller than 1.3 m, which involve undetectable small or understory trees. Furthermore, we demonstrate that only marginal effort is needed to transfer our framework to data from Finland, despite markedly dissimilar data sources. Our work lays the foundation for digitalized national databases, where large trees are spatially traceable and manageable.

Estimation of carbon sequestration by some selected dominant tree species of Chandigarh University, Gharuan, Mohali

The purpose of this study was to determine the carbon in selected tree species at Chandigarh University to analysis carbon stock by measuring their aboveground and belowground biomass for sequestering carbon. Distinct Count's tree species have a total biomass of 113. There are a total of 701 trees of 54 different kinds growing in a particular region of the Chandigarh University campus in Punjab. Non-destructive methods have been used to measure the total biomass and carbon sequestered by the tree species. Each species' above and belowground organic carbon (tones/tree) and total organic carbon was determined. The overall total of carbon that may be found in different tree species is 17474.22.

Evaluation of species-specific tree density and height interpreted from airborne digital stereo imagery in young boreal stands in Northwestern Ontario

Monitoring silviculture through accurate assessment of the density and height of trees in young (10–25 years) stands is a critical component of sustainable forest management. Reporting species composition and height of young stands that regenerate after harvest and renewal treatments ensures planned targets have been achieved. In the boreal regions of Ontario, operational assessment of young stand conditions is conducted using qualitative visual and/or higher cost quantitative plot-based field assessments. High resolution three-dimensional digital imagery can be collected using manned aircraft across large forest management areas for stereo-photo interpretation. Our objective was to evaluate the accuracy of stereo-photo interpretation of species-specific tree counts and height in plots in digital imagery, and the effect of resolution on interpretation accuracy. Digital imagery (7-cm and 10-cm resolution) was acquired across nine stands representing common regeneration types. Prior to imagery acquisition, marked plots were established and assessed in the field; 177 plots were used in the analysis. Higher resolution imagery improved accuracy of total and conifer tree counts and conifer heights. Interpreted counts of white birch trees and height estimates of deciduous stems were not accurate using the leaf-off imagery, and interpreters could not differentiate spruce from fir.

Leveraging on Advanced Remote Sensing- and Artificial Intelligence-Based Technologies to Manage Palm Oil Plantation for Current Global Scenario: A Review

Advanced remote sensing technologies have undoubtedly revolutionized palm oil industry management by bringing business and environmental benefits on a single platform. It is evident from the ongoing trend that remote sensing using satellite and aerial data is able to provide precise and quick information for huge palm oil plantation areas using high-resolution image processing, which is also recognized by the certification agencies, i.e., the Roundtable on Sustainable Palm Oil (RSPO) and ISCC (International Sustainability and Carbon Certification). A substantial improvement in the palm oil industry could be attained by utilizing the latest Geo-information tools and technologies equipped with AI (Artificial Intelligence) algorithms and image processing, which could help to identify illegal deforestation, tree count, tree height, and the early detection of diseased leaves. This paper reviews some of the latest technologies equipped with remote sensing, AI, and image processing for managing the palm oil plantation. This manuscript also highlights how the distress in the current palm oil industry could be handled by mentioning some of the improvised monitoring systems for palm oil plantation that could in turn increase the yield of palm oil. It is evident from the proposed review that the accuracy of AI algorithms for palm oil detection depends on various factors such as the quality of the training data, the design of the neural network, and the type of detection task. In general, AI models have achieved high accuracy in detecting palm oil tree images, with some studies reporting accuracy levels up to 91%. However, it is important to note that accuracy can still be affected by factors such as variations in lighting conditions and image resolution. Nonetheless, with any AI model, the accuracy of algorithms for palm oil tree detection can be improved by collecting more diverse training data and fine-tuning the model.

Crown-Level Structure and Fuel Load Characterization from Airborne and Terrestrial Laser Scanning in a Longleaf Pine (Pinus palustris Mill.) Forest Ecosystem

Airborne Laser Scanners (ALS) and Terrestrial Laser Scanners (TLS) are two lidar systems frequently used for remote sensing forested ecosystems. The aim of this study was to compare crown metrics derived from TLS, ALS, and a combination of both for describing the crown structure and fuel attributes of longleaf pine (Pinus palustris Mill.) dominated forest located at Eglin Air Force Base (AFB), Florida, USA. The study landscape was characterized by an ALS and TLS data collection along with field measurements within three large (1963 m2 each) plots in total, each one representing a distinct stand condition at Eglin AFB. Tree-level measurements included bole diameter at breast height (DBH), total height (HT), crown base height (CBH), and crown width (CW). In addition, the crown structure and fuel metrics foliage biomass (FB), stem branches biomass (SB), crown biomass (CB), and crown bulk density (CBD) were calculated using allometric equations. Canopy Height Models (CHM) were created from ALS and TLS point clouds separately and by combining them (ALS + TLS). Individual trees were extracted, and crown-level metrics were computed from the three lidar-derived datasets and used to train random forest (RF) models. The results of the individual tree detection showed successful estimation of tree count from all lidar-derived datasets, with marginal errors ranging from −4 to 3%. For all three lidar-derived datasets, the RF models accurately predicted all tree-level attributes. Overall, we found strong positive correlations between model predictions and observed values (R2 between 0.80 and 0.98), low to moderate errors (RMSE% between 4.56 and 50.99%), and low biases (between 0.03% and −2.86%). The highest R2 using ALS data was achieved predicting CBH (R2 = 0.98), while for TLS and ALS + TLS, the highest R2 was observed predicting HT, CW, and CBD (R2 = 0.94) and HT (R2 = 0.98), respectively. Relative RMSE was lowest for HT using three lidar datasets (ALS = 4.83%, TLS = 7.22%, and ALS + TLS = 4.56%). All models and datasets had similar accuracies in terms of bias (<2.0%), except for CB in ALS (−2.53%) and ALS + TLS (−2.86%), and SB in ALS + TLS data (−2.22%). These results demonstrate the usefulness of all three lidar-related methodologies and lidar modeling overall, along with lidar applicability in the estimation of crown structure and fuel attributes of longleaf pine forest ecosystems. Given that TLS measurements are less practical and more expensive, our comparison suggests that ALS measurements are still reasonable for many applications, and its usefulness is justified. This novel tree-level analysis and its respective results contribute to lidar-based planning of forest structure and fuel management.

Individual tree segmentation and tree-counting using supervised clustering

Individual tree segmentation (ITS), or tree counting, is a fundamental work in precision forestry and agriculture processes. Rather than the time-consuming and labor-intensive manual inspection, computer vision has shown substantial prospects in unmanned aerial vehicle (UAV)-based applications; one such application includes the automatic tree-counting problem in the forest resource inventory. However, it is difficult to obtain individual trees owing to the particularity of tree canopy crowns, such as color graduality, shape uncertainty, and overlapping. In this study, we propose a learning framework based on supervised data clustering. Here, both ITS and tree counting can be obtained accurately and simultaneously from a two-dimensional (2D) top-view tree canopy crowns whether they are isolated, overlapped, or both. A pixel-precision classifier is used to recognize tree pixels or superpixels (a set of meaningful pixels), rather than using complex image preprocessing or feature construction techniques. The obtained tree superpixels are then grouped into individual trees by the proposed supervised clustering method. To obtain accurate ITS and tree counts, the similarity used for the clustering is “learned” from the user-supplied supervisions, rather than “pre-specified” or “grid-search” as in the existing state-of-the-art methods. This study also includes an extensive experimental comparison of homogenous and heterogeneous high-resolution images. The results demonstrate that our method is superior to the state-of-the-art methods in both visualized ITS and numeric tree-counting results, even comparable to human vision. It achieves a counting accuracy of 99.16 % in terms of the R2 value and 2.2923 in terms of the pixel mean absolute error (MAE). These are 22.04 % higher and 8.727 lower, respectively, than those of the second-best method.

Tree Seedlings Detection and Counting Using a Deep Learning Algorithm

Tree-counting methods based on computer vision technologies are low-cost and efficient in contrast to the traditional tree counting methods, which are time-consuming, laborious, and humanly infeasible. This study presents a method for detecting and counting tree seedlings in images using a deep learning algorithm with a high economic value and broad application prospects in detecting the type and quantity of tree seedlings. The dataset was built with three types of tree seedlings: dragon spruce, black chokeberries, and Scots pine. The data were augmented via several data augmentation methods to improve the accuracy of the detection model and prevent overfitting. Then a YOLOv5 object detection network was built and trained with three types of tree seedlings to obtain the training weights. The results of the experiments showed that our proposed method could effectively identify and count the tree seedlings in an image. Specifically, the MAP of the dragon spruce, black chokeberries, and Scots pine tree seedlings were 89.8%, 89.1%, and 95.6%, respectively. The accuracy of the detection model reached 95.10% on average (98.58% for dragon spruce, 91.62% for black chokeberries, and 95.11% for Scots pine). The proposed method can provide technical support for the statistical tasks of counting trees.

Stand validation of lidar forest inventory modeling for a managed southern pine forest

We evaluated area-based approaches (ABAs) to light detection and ranging (lidar) predictions of plot- and stand-level forest attributes (tree count, height, basal area, volume, aboveground biomass, broadleaf/conifer, and diameter at breast height — “diameter”). ABA methods included post-stratification (PS), ordinary least squares (OLSs) regression, k nearest neighbors ( kNN), and random forest (RF). This study was conducted on the Savannah River Site in South Carolina, USA. Plot- and stand-level predictions were validated against fixed-radius 0.04 ha (0.1 acre) plots in 49 ≈2.0 ha (5 acre) stands. Our findings demonstrate that lidar can be incorporated operationally into forest inventory systems to provide stand-level inferences for a wide range of forest attributes. Volume predictions for specific diameter classes, however, often fared poorly (root mean squared error (RMSE) > 100%) for the methods we explored, especially for larger (less common) diameter trees. Stand-level results were consistently better than pixel-level results (10–200+ percentage points). kNN and RF performed similarly and better than OLS and PS, but RF was the most robust to model configurations, while kNN has practical advantages such as simultaneous predictions of many attributes.

Soil and Vegetation Heterogeneity in Postagrogenic Restorative Succession in the Middle Taiga Subzone

The specific features of the soil–vegetation heterogeneity at the early stages of restorative succession have been studied. These features are determined by biotic and bioinert conditions that were influenced by anthropogenic factors during long-term plowing on azonal carbonated soils in the middle taiga subzone (Kargopol district, Arkhangelsk Oblast, Russia). In total, 25 full-profile soil sections, 500 soil samples, 125 sites for counting tree and shrub species, 25 standard geobotanical plots, and 250 plots for counting herbaceous species have been studied and analyzed. During the first 15 years of restorative succession on fallow agricultural land, herbaceous vegetation has the highest phytocenotic significance. On the residual calcareous soils of the forest zone, the specialization of ecological properties, i.e., the confinement of species to certain factors is weakly pronounced only in some herbaceous species. The mosaic structure of vegetation cover on a 15-yearold fallow is determined not by the variability of soil properties and the influence of growing tree and shrub species, but by the biological properties of herbaceous plants and their specific requirements for environmental factors. Soil fertility and soil moisture are the main factors that determine the variations of vegetation at this stage.

Nation-wide mapping of tree-level aboveground carbon stocks in Rwanda

Trees sustain livelihoods and mitigate climate change but a predominance of trees outside forests and limited resources make it difficult for many tropical countries to conduct automated nation-wide inventories. Here, we propose an approach to map the carbon stock of each individual overstory tree at the national scale of Rwanda using aerial imagery from 2008 and deep learning. We show that 72% of the mapped trees are located in farmlands and savannas and 17% in plantations, accounting for 48.6% of the national aboveground carbon stocks. Natural forests cover 11% of the total tree count and 51.4% of the national carbon stocks, with an overall carbon stock uncertainty of 16.9%. The mapping of all trees allows partitioning to any landscapes classification and is urgently needed for effective planning and monitoring of restoration activities as well as for optimization of carbon sequestration, biodiversity and economic benefits of trees.

Automatic detection and counting of oil palm trees using remote sensing and object-based deep learning

Palm oil is one of the world's highest vegetable oil producers. Providing accurate oil palm plantation statistics is essential to supporting effective and efficient decision-making, especially since the oil palm industry has a strategic role in achieving several Sustainable Development Goals (SDGs). The existing conventional data collection has been focusing on field survey methods which require considerable human resources and costs, long collection and processing times, and have difficulty in reaching remote areas. Remote sensing with satellite imagery and Unmanned Aerial Vehicles (UAV) can be an alternative due to its advantages which include a more efficient workforce, shorter update times, comprehensive area coverage, and the ability to reach remote locations. In this study, we propose the utilization of Microsoft Bing Maps Very High Resolution (VHR) satellite imagery and UAV with image processing threshold and object-based deep learning methods to detect and count the number of oil palm trees. Experimental results in a case study in Rokan Hulu Regency, Riau, Indonesia show that the object-based deep learning model of the You Only Look Once (YOLO) architecture achieves excellent performance with an F1-Score of up to 91.05%. This model is superior to the widely adopted baseline model of the image processing threshold, which only reached an F1-Score of 40.59%. Our findings suggest the use of very high resolution images with object-based deep learning is promising for automatic detection and counting of oil palm trees to support the accurate and sustainable agricultural monitoring.

The Role of Remote Sensing in Olive Growing Farm Management: A Research Outlook from 2000 to the Present in the Framework of Precision Agriculture Applications

Given the importance of olive growing, especially in Mediterranean countries, it is crucial that there is a constant process of modernization aimed at both environmental sustainability and the maintenance of high standards of production. The use of remote sensing (RS) allows intervention in a specific and differentiated way in olive groves, depending on their variability, in managing different agronomic aspects. The potentialities of the application of RS in olive growing are topics of great agronomic interest to olive growers. Using the tools provided by RS and the modernization of the olive sector can bring great future prospects by reducing costs, optimizing agronomic management, and improving production quantity and quality. This article is part of a review that aims to cover the past, from the 2000s onwards, and the most recent applications of aerial RS in olive growing in order to be able to include research and all topics related to the use of RS on olive trees. As far as the use of RS platforms such as satellites, aircraft, and unmanned aerial vehicles (UAVs) as olive growing is concerned, a literature review showed the presence of several works devoted to this topic. This article covers purely agronomic matters of interest to olive farms (and related research that includes the application of RS), such as yielding and managing diseases and pests, and detection and counting of olive trees. In addition to these topics, there are other relevant aspects concerning the characterization of the canopy structure of olive trees which is particularly interesting for mechanized pruning management and phenotyping.

The taxicab sampler: MCMC for discrete spaces with application to tree models

Motivated by the problem of exploring discrete but very complex state spaces in Bayesian models, we propose a novel Markov Chain Monte Carlo search algorithm: the taxicab sampler. We describe the construction of this sampler and discuss how its interpretation and usage differs from that of standard Metropolis-Hastings as well as the related Hamming ball sampler. The proposed sampling algorithm is then shown to demonstrate substantial improvement in computation time without any loss of efficiency relative to a naïve Metropolis–Hastings search in a motivating Bayesian regression tree count model, in which we leverage the discrete state space assumption to construct a novel likelihood function that allows for flexibly describing different mean-variance relationships while preserving parameter interpretability compared to existing likelihood functions for count data.

Reviewed Commentary: Factory-Grown Wood, the Future of Forestry?

Recent developments in factory-grown foods suggest that factory-grown wood (FGW) may be on the horizon. In fact, recent work at Massachusetts Institute of Technology introduces tunable plant-based materials, an early indicator of what may evolve into a new source of raw material for forest sector companies, and others. Industry and academia would be wise to monitor developments in this field as they may present significant opportunities and/or adjustments for both. We explore the state-of-the-art in this budding area of science and contemplate implications of successfully growing wood or other lignocellulosic materials in factories. Given a changing climate and focus on carbon emissions, the pressure to drastically reduce CO2 production will continue climb. Could reduction of their footprint via FGW be an important part of this equation for forest sector companies, going beyond the need to “make every tree count”? In other words, might FGW present an environmental and climate protection breakthrough? Or might it simply trade forest-based environmental impacts for others? What other consequences does FGW promise for companies? And, what might it mean for wood science programs, critical suppliers of research & development and skilled employees for the industry? We explore each of these questions and contemplate potential actions and outcomes.