Multisensor Data Fusion for Improved Segmentation of Individual Tree Crowns in Dense Tropical Forests

Aubry‐Kientz, Mélaine; Laybros, Anthony; Ball, J.G.; Jackson, Toby; Coomes, David A.; Vincent, Grégoire

doi:10.1109/jstars.2021.3069159

Cited by 24 publications

(15 citation statements)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The study site is Paracou (5 • 18'N, 52 • 55'W) in French Guiana (FG), which is a tropical rain forest. The data set comprises 374 bands, 76 tree species with 1 m spatial resolution [19] (consisting of 668 × 923 pixels). The ground truth labels were manually confirmed from field survey and the crowns manually segmented using using the QGIS open-source software.…”

Section: Experiments and Resultsmentioning

confidence: 99%

Tree species classification from hyperspectral data using graph-regularized neural networks

Bandyopadhyay¹,

Mukherjee²

2022

Preprint

View full text Add to dashboard Cite

Manual labeling of tree species remains a challenging task, especially in tropical regions, owing to inaccessibility and labor-intensive ground-based surveys. Hyperspectral images (HSIs), through their narrow and contiguous bands, can assist in distinguishing tree species based on their spectral properties. Therefore, automated classification algorithms on HSI images can help augment the limited labeled information and generate a real-time classification map for various tree species. Achieving high classification accuracy with a limited amount of labeled information in an image is one of the key challenges that researchers have started addressing in recent years. We propose a novel graph-regularized neural network (GRNN) algorithm that encompasses the superpixel-based segmentation for graph construction, a pixel-wise neural network classifier, and the label propagation technique to generate an accurate classification map. GRNN outperforms several state-of-the-art techniques not only for the standard Indian Pines HSI but also achieves a high classification accuracy (approx. 92%) on a new HSI data set collected over the forests of French Guiana (FG) even when less than 1% of the pixels are labeled. We show that GRNN is not only competitive with the state-of-the-art semi-supervised methods, but also exhibits lower variance in accuracy for different number of training samples and over different independent random sampling of the labeled pixels for training.

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Tree species classification from hyperspectral data using graph-regularized neural networks

Bandyopadhyay¹,

Mukherjee²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Hyperspectral data, especially in the infrared spectrum, is often used for differentiating tree species based on spectral differences among species in leaf chemistry and canopy structure [ 22 ]. Hyperspectral data is particularly useful in forests with high species diversity where neighboring trees are likely to be different species and thus spectrally distinct ( Fig 3 )[ 23 ]. All hyperspectral data were collected during the same field collection campaign as the RGB data, with the exception of the UNDE site, in which the 2019 RGB data was not available at the time of publication and therefore the 2017 flight data was used instead.…”

Section: Methods and Resultsmentioning

confidence: 99%

“…To demonstrate the performance of a detection method on the benchmark dataset and allow for users to gauge their performance against published methods, we used the DeepForest Python package to generate crown detections in the benchmark sensor data [ 34 ]. DeepForest is a RGB deep learning model that predicts canopy crown bounding boxes [ 11 , 23 , 35 ]. The prebuilt model in DeepForest was trained with the training data described above, but did not use or overlap spatially with any evaluation data in this benchmark.…”

Section: Methods and Resultsmentioning

confidence: 99%

A benchmark dataset for canopy crown detection and delineation in co-registered airborne RGB, LiDAR and hyperspectral imagery from the National Ecological Observation Network

et al. 2021

View full text Add to dashboard Cite

Broad scale remote sensing promises to build forest inventories at unprecedented scales. A crucial step in this process is to associate sensor data into individual crowns. While dozens of crown detection algorithms have been proposed, their performance is typically not compared based on standard data or evaluation metrics. There is a need for a benchmark dataset to minimize differences in reported results as well as support evaluation of algorithms across a broad range of forest types. Combining RGB, LiDAR and hyperspectral sensor data from the USA National Ecological Observatory Network’s Airborne Observation Platform with multiple types of evaluation data, we created a benchmark dataset to assess crown detection and delineation methods for canopy trees covering dominant forest types in the United States. This benchmark dataset includes an R package to standardize evaluation metrics and simplify comparisons between methods. The benchmark dataset contains over 6,000 image-annotated crowns, 400 field-annotated crowns, and 3,000 canopy stem points from a wide range of forest types. In addition, we include over 10,000 training crowns for optional use. We discuss the different evaluation data sources and assess the accuracy of the image-annotated crowns by comparing annotations among multiple annotators as well as overlapping field-annotated crowns. We provide an example submission and score for an open-source algorithm that can serve as a baseline for future methods.

show abstract

“…Delineating individual trees from airborne lidar datasets is most successful for coniferous forests, because their apical dominance results in clear local height maxima that make tree crowns easily distinguishable (11, 23), but complex tropical canopies have presented a far greater challenge for lidar delineation (24). Tropical forest canopies are often densely packed with partially interwoven trees which point-cloud clustering algorithms can struggle to distinguish (25). Furthermore, airplanes or helicopters are required to conduct lidar surveys whereas standard RGB imagery is far cheaper and easier to collect because of the availability of low-cost drones.…”

Section: Introductionmentioning

confidence: 99%

Accurate delineation of individual tree crowns in tropical forests from aerial RGB imagery using Mask R-CNN

Hickman¹,

Ball

Jackson

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Tropical forests are a major component of the global carbon cycle and home to two-thirds of terrestrial species. Upper-canopy trees store the majority of forest carbon and can be particularly vulnerable to drought events and storms. Monitoring their growth and mortality is essential to understanding forest resilience to climate change, but large trees are underrepresented in traditional field surveys, so estimates are poorly constrained. Aerial photographs provide spectral and textural information to discriminate between tree crowns in diverse, complex tropical canopies, potentially opening the door to landscape monitoring of large trees. Here we describe a new deep convolutional neural network machine learning method, Detectree2, which builds on the Mask R-CNN computer vision framework to recognise the irregular edges of individual tree crowns from airborne RGB imagery. We trained and evaluated this model with 3,600 manually delineated tree crowns at three sites in Malaysian Borneo and one site in French Guiana. As an example application of this tool, we combined the delineations with repeat lidar surveys of the four sites to estimate the growth and mortality of upper-canopy trees. Detectree2 delineated 65,000 upper-canopy trees across 14 km^2 of aerial images. The skill of the automatic method in delineating the upper-canopy tree crowns was good (F1 = 0.71). As predicted from previous field studies, we found that growth rate declined with tree height and tall trees had higher mortality rates than intermediate-size trees. In addition, tall trees in French Guiana had higher growth and mortality rates than those in Borneo. Our approach demonstrates that machine learning methods can automatically segment trees in abundant, cheap RGB imagery. This tool has many potential applications in forest ecology and conservation, from estimating carbon stocks to monitoring forest restoration. We demonstrate its use in tracking the growth and mortality rates of upper-canopy trees at scales much larger than those achievable with field data.

show abstract

Multisensor Data Fusion for Improved Segmentation of Individual Tree Crowns in Dense Tropical Forests

Cited by 24 publications

References 64 publications

Tree species classification from hyperspectral data using graph-regularized neural networks

Tree species classification from hyperspectral data using graph-regularized neural networks

A benchmark dataset for canopy crown detection and delineation in co-registered airborne RGB, LiDAR and hyperspectral imagery from the National Ecological Observation Network

Accurate delineation of individual tree crowns in tropical forests from aerial RGB imagery using Mask R-CNN

Contact Info

Product

Resources

About