Skip to main content

LUP Student Papers

LUND UNIVERSITY LIBRARIES

Urban tree canopy mapping -an open source deep learning approach

Rydlinge, Tobias LU (2023) In Master Thesis in Geographical Information Science GISM01 20231
Dept of Physical Geography and Ecosystem Science
Abstract
Urban trees have an important role to provide ecosystem services and to make our cities greener and more sustainable. The changing climate and densification of cities make it even more valuable to preserve and investigate in urban trees. Tree canopy detection in cities is challenging, with both trees and other objects of irregular shape, size and complexity. The aerial images from Lantmäteriet (the Swedish mapping cadastral and land registration authority) is a great source for image analysis on high resolution images, but the leaf-off images may be challenging for tree canopy detection.
The aim of the study was to test if Light Detection and Ranging (LIDAR) could be combined with aerial images to develop a deep learning tool for urban... (More)
Urban trees have an important role to provide ecosystem services and to make our cities greener and more sustainable. The changing climate and densification of cities make it even more valuable to preserve and investigate in urban trees. Tree canopy detection in cities is challenging, with both trees and other objects of irregular shape, size and complexity. The aerial images from Lantmäteriet (the Swedish mapping cadastral and land registration authority) is a great source for image analysis on high resolution images, but the leaf-off images may be challenging for tree canopy detection.
The aim of the study was to test if Light Detection and Ranging (LIDAR) could be combined with aerial images to develop a deep learning tool for urban tree canopy mapping under leaf-off conditions. The deep learning method using LIDAR and aerial leaf-off images had a precision of 88 % mapping urban tree canopy. Using the same method with LIDAR and Infrared (IR) leaf-on data yielded a precision of 91 %. The LIDAR data increases the accuracy for leaf-off data when added to the deep learning model.
The findings of this study indicate that tree canopy mapping with LIDAR and aerial images taken under leaf-off conditions can be used for tree canopy mapping with comparable results to other methods. (Less)
Popular Abstract (Swedish)
Träd är viktiga kolsänkor och bildar vår världs gröna lungor. De utför viktiga ekosystemtjänster i våra städer och har tagit en allt viktigare roll i våra stadsrum. Utöver detta ökar närheten till träd vårt välbefinnande och vår fysiska och psykiska hälsa.
Medborgarnas närhet till träd och tätortsskog blir något viktigt för alla städer och därmed sätts också nyttan av att kartlägga den nuvarande geografiska utbredningen av träd i fokus. Var bör vi plantera mer träd? Var ska träd planteras för motverka värmeöar från solstrålning? Beror skillnader i närhet till träd på vilket område vi bor i?
Maskininlärning med öppen källkod på data som kommunerna själva har tillgång till gör denna metod tillämpbar och reproducerbar. Med... (More)
Träd är viktiga kolsänkor och bildar vår världs gröna lungor. De utför viktiga ekosystemtjänster i våra städer och har tagit en allt viktigare roll i våra stadsrum. Utöver detta ökar närheten till träd vårt välbefinnande och vår fysiska och psykiska hälsa.
Medborgarnas närhet till träd och tätortsskog blir något viktigt för alla städer och därmed sätts också nyttan av att kartlägga den nuvarande geografiska utbredningen av träd i fokus. Var bör vi plantera mer träd? Var ska träd planteras för motverka värmeöar från solstrålning? Beror skillnader i närhet till träd på vilket område vi bor i?
Maskininlärning med öppen källkod på data som kommunerna själva har tillgång till gör denna metod tillämpbar och reproducerbar. Med trädkartläggningen kan datadrivna beslut fattas baserat på en uppdaterad och detaljerad trädkartläggning. Detta underlättar också kostnadseffektivt underhåll av urbana träd, genom att till exempel upprätthålla en sammanhållen trädstruktur.
Syftet med studien var att testa om Light Detection and Ranging (LIDAR) kan kombineras med flygbilder för att utveckla en maskininlärningsmodell för kartläggning av stadsträd. Metoden hade en precision på 88 %. Att använda samma metod med LIDAR och lövade Infraröda (IR) bilder gav en precision på 91 %.
Resultaten indikerar att LIDAR och flygbilder tagna under avlövade förhållanden kan användas för kartläggning av stadsträd med jämförbara resultat med andra metoder. (Less)
Popular Abstract
Trees are important carbon sinks and form the green lungs of our world. They perform important ecosystem services in our cities and have taken an increasingly important role in our urban spaces. In addition to this, proximity to trees increases our well-being and our physical and mental health.
Citizen proximity to trees and urban forest becomes something important for all cities, and thus the benefit of mapping the current geographical distribution of trees is also brought into focus. Where are the plantable spots? Where should trees be planted to defeat urban heat islands from solar radiation? Do differences of tree proximity depend on which area you live in?
Machine learning with open source code on data that the municipalities... (More)
Trees are important carbon sinks and form the green lungs of our world. They perform important ecosystem services in our cities and have taken an increasingly important role in our urban spaces. In addition to this, proximity to trees increases our well-being and our physical and mental health.
Citizen proximity to trees and urban forest becomes something important for all cities, and thus the benefit of mapping the current geographical distribution of trees is also brought into focus. Where are the plantable spots? Where should trees be planted to defeat urban heat islands from solar radiation? Do differences of tree proximity depend on which area you live in?
Machine learning with open source code on data that the municipalities themselves have access to makes the method of this thesis applicable and reproducible. With the tree mapping, data-driven decisions can be made based on updated and detailed data. This also facilitates cost-effective maintenance of urban trees, for example by maintaining a cohesive tree structure.
The aim of the study was to test if Light Detection and Ranging (LIDAR) could be combined with aerial images to develop a deep learning tool for urban tree canopy mapping under leaf-off conditions. The deep learning method using LIDAR and aerial leaf-off images had a precision of 88 % mapping urban tree canopy. Using the same method with LIDAR and Infrared (IR) leaf-on data yielded a precision of 91 %.
The findings of this study indicate that tree canopy mapping with LIDAR and aerial images taken under leaf-off conditions can be used for tree canopy mapping with comparable results to other methods. (Less)
Please use this url to cite or link to this publication:
author
Rydlinge, Tobias LU
supervisor
organization
course
GISM01 20231
year
type
H2 - Master's Degree (Two Years)
subject
keywords
Geography, GIS, AI, deep learning, urban trees, remote sensing
publication/series
Master Thesis in Geographical Information Science
report number
157
language
English
id
9112048
date added to LUP
2023-03-09 09:43:13
date last changed
2023-03-09 09:43:13
@misc{9112048,
  abstract     = {{Urban trees have an important role to provide ecosystem services and to make our cities greener and more sustainable. The changing climate and densification of cities make it even more valuable to preserve and investigate in urban trees. Tree canopy detection in cities is challenging, with both trees and other objects of irregular shape, size and complexity. The aerial images from Lantmäteriet (the Swedish mapping cadastral and land registration authority) is a great source for image analysis on high resolution images, but the leaf-off images may be challenging for tree canopy detection.
The aim of the study was to test if Light Detection and Ranging (LIDAR) could be combined with aerial images to develop a deep learning tool for urban tree canopy mapping under leaf-off conditions. The deep learning method using LIDAR and aerial leaf-off images had a precision of 88 % mapping urban tree canopy. Using the same method with LIDAR and Infrared (IR) leaf-on data yielded a precision of 91 %. The LIDAR data increases the accuracy for leaf-off data when added to the deep learning model.
The findings of this study indicate that tree canopy mapping with LIDAR and aerial images taken under leaf-off conditions can be used for tree canopy mapping with comparable results to other methods.}},
  author       = {{Rydlinge, Tobias}},
  language     = {{eng}},
  note         = {{Student Paper}},
  series       = {{Master Thesis in Geographical Information Science}},
  title        = {{Urban tree canopy mapping -an open source deep learning approach}},
  year         = {{2023}},
}