Narrow your search

Library

ULiège (2)


Resource type

dissertation (2)


Language

English (2)


Year
From To Submit

2022 (1)

2021 (1)

Listing 1 - 2 of 2
Sort by

Dissertation
Assessment of Sitka spruce natural regeneration and recruitment in Scotland through photogrammetry
Authors: --- --- --- --- --- et al.
Year: 2021 Publisher: Liège Université de Liège (ULiège)

Loading...
Export citation

Choose an application

Bookmark

Abstract

In Scotland, Sitka spruce represents the major timber resource. The regeneration of this species isabundant and occurs widely after clearfelling. Finding way to carry out affordable inventories of
regeneration would help forest managers to make decision on the future of the regeneration resource.
The deployment of Unmanned Aerial Vehicle (UAV) allows rapid assessment of forest and regenerationand it is likely to lead of a decreasing cost for field surveys.
This current study aims to map Sitka spruce seedlings with orthophotos acquired by drones. The used method, based on RGB images, has been carried out using OTB (Orfeo ToolBox) and a object based image analysis approach. It consists in four steps: the segmentation (i), a supervised classification (random forest) (ii), the mapping of the studied sites (iii) and finally, the validation (iv) of the model through points acquired by photointerpretation. The result shows that one of the classification models reaches a global accuracy of 66.9 with pseudo independent dataset and 77.4 with an independent dataset. The results are expected to be better with images acquired during other periods (leaf-off period) in order to prevent confusion with surrounding vegetation presents on the studied sites. Despite this fact, the mapping of Sitka spruce seems promising with an RGB camera and may offer a promising potential for commercial forestry. In addition, the method may be applied in other context such as ecological restoration or forest health. En Ecosse, l'épicéa de Sitka représente la principale ressource en bois. La régénération de cette espéce est abondante, particulièrement aprés une coupe à blanc. Trouver un moyen de réaliser des inventaires abordables de celle-ci aiderait les gestionnaires forestiers à prendre des décisions sur l'avenir de la ressource. L'utilisation de drones permet une évaluation rapide de la forêt et de la régénération, permettant ainsi une réduction des coûts des inventaires de terrain.
La présente étude vise à cartographier les semis d'épicéa de Sitka à partir d'orthophotos acquises par drones. La méthode utilisée, basée sur des images RGB, a été réalisée en utilisant OTB (Orfeo ToolBox) et une approche d'analyse d'image basée sur des objets. Elle consiste en quatre étapes: la segmentation (i), une classification supervisée (random forest) (ii), la cartographie des sites étudiés(iii) et enfin, la validation (iv) du modéle par des points acquis par photointerprétation. Les résultats montrent qu'un des modéles de classification atteint une précision globale de 66,9 avec un jeu de données pseudo-indépendant et de 77,4 avec un jeu de données indépendant. On s'attend à ce que les résultats soient meilleurs avec des images acquises à d'autres périodes de l'années (période hors feuille) afin d'éviter la confusion avec la végétation environnante présentes sur les sites d'étude. Malgré cela, la cartographie de l'épicéa de Sitka semble prometteuse avec une caméra RGB et offrirait donc un potentiel intéressant pour la foresterie commerciale. En outre, la méthode pourrait être utilisée dans d'autres contextes, tel que celui de la restauration écologique ou la santé des forêts.


Dissertation
Regional scale mapping and characterization of the poplar resource using two remote sensing data mining approaches
Authors: --- --- --- --- --- et al.
Year: 2022 Publisher: Liège Université de Liège (ULiège)

Loading...
Export citation

Choose an application

Bookmark

Abstract

In Wallonia, the regional forest resources are estimated from field methods that may present biases for estimating fast-growing forest species area like poplars, thus requiring support from remote sensing-based solutions. 
The objectives of this master thesis concern the mapping and characterization of the poplar resource in the province of Hainaut. Specifically, it investigates (i) the potential of S2 super-resolution images and (ii) the use of orthoimages through a deep learning-based approach to map the poplar resource, followed by (iii) the ability of an aerial photogrammetry CHM to characterize the latter.
The used methods are divided into two approaches: classification of super-resolved s2 images using Random Forest algorithm (Breiman, 2001), semantic segmentation of ortho-images through a Deep Layer Aggregation (Yu et al., 2018) Neural Network. Both approaches involve 5 steps: data preparation, supervised learning, map production, height classification and accuracy assessment.
The results for the first approach map, with a F1-score of 0.923, is limited in detecting young poplar plantations and overestimates the poplar resource. Then, the second approach produced a map presenting great potential to detect poplar trees with an average accuracy of 1m between the position of correctly predicted and observed poplars, but still contains many False Negatives, resulting in a F1-score of 0.653. Finally, poplar resource characterization shows for the first and second approach a respective ratio of properly identified height classes of 50% and 69%, these results are contrasted by poor ground truth data and a convincing visual assessment.
To conclude, the super-resolution of sentinel-2 image seems to bring a higher accuracy compared to the poplar resource map made on S2 images by (Bolyn, Latte, Colson, et al., 2020a). Furthermore, a potential to map the poplar resource from orthoimages using a deep learning-based approach has been highlighted in this project, despite a low accuracy to be the subject of a management tool at this time. Lastly, although contrasting results, it would seem that aerial photogrammetry CHM could be appropriate to characterize the poplar resource in this project, but would require field validation.

Listing 1 - 2 of 2
Sort by