Faculté des Sciences appliquées
Faculté des Sciences appliquées

Change Detection on Sentinel-2 multi-spectral images: A Semi-(Non-)Supervised Learning approach

Pirenne, Thomas ULiège
Promotor(s) : Geurts, Pierre ULiège ; Draime, Damien
Date of defense : 24-Jun-2021/25-Jun-2021 • Permalink :
Title : Change Detection on Sentinel-2 multi-spectral images: A Semi-(Non-)Supervised Learning approach
Translated title : [fr] Détection de Changements dans les images multispectrales de Sentinel-2: Une approche d'Apprentissage Semi-(Non-)Supervisée
Author : Pirenne, Thomas ULiège
Date of defense  : 24-Jun-2021/25-Jun-2021
Advisor(s) : Geurts, Pierre ULiège
Draime, Damien 
Committee's member(s) : Van Droogenbroeck, Marc ULiège
Wehenkel, Louis ULiège
Language : English
Number of pages : 90-100
Keywords : [en] Multispectral Imagery
[en] Change Detection
[en] Unsupervised Learning
[en] Semi-Supervised Learning
[en] Generative Adversarial Networks
[en] Deep Learning
[en] Machine Learning
[en] Deep Convolutional Generative Adversarial Networks
[en] High Resolution Satellite Imagery
[en] Sentinel-2
Discipline(s) : Engineering, computing & technology > Computer science
Commentary : This thesis has been carried out in the context of an internship at Aerospacelab.
Target public : Professionals of domain
Institution(s) : Université de Liège, Liège, Belgique
Degree: Master en ingénieur civil en informatique, à finalité spécialisée en "intelligent systems"
Faculty: Master thesis of the Faculté des Sciences appliquées


[en] Land cover change detection is a crucial task to automate for many applications ranging from efficient natural disaster monitoring or military surveillance to property insurance optimization. Space missions, notably Sentinel-2 from the European Space Agency have opened the path to a much wider use of multispectral geospatial imaging with high temporal resolution which proves ideal for the task of bitemporal change detection. In that context, the company Aerospacelab has implemented a neural network trained on a synthetic dataset in order to cope for the lack of labeled data. This thesis has two aims: evaluating the potential of multispectral channels in contrast to the three typical visual spectral channels Red-Green-Blue for the task of general change detection and establishing whether or not unsupervised or semi-supervised methods can better cope with the common lack of available labeled data.

To provide these insights, an unsupervised approach initially proposed by Gong et al. called Generative Discriminatory Classified Network (GDCN) is exploited, derived, implemented and tested. It consists in a generator to produce bitemporal multispectral satellite images and a discriminator which simultaneously discriminates between real and fake pairs of images as well as classify each of its pixels as change or non-change. To determine the impact of each of the approach's components on the performances for the task of change detection, three models are implemented: GDCN itself, ConvGDCN which is a scalable derivation of the model inspired from the Deep Convolutional GAN framework proposed by Radford et al. and finally, ConvCN which is a copy of ConvGDCN from which were removed the parts dedicated to the generation task in order to evaluate just how useful it is. Each model is trained in the unsupervised manner proposed by Gong et al. relying on pseudo labeled samples provided by another notable unsupervised method: Compressed Change Vector Analysis by Bovolo et al. The unsupervised models are tuned on a validation set and the best of each model trained on both RGB and multispectral images are fine-tuned with a small amount of hand-labeled samples.

All unsupervised and tuned models with both RGB and multispectral images are tested on a test dataset and the results provide insights to the company's questions. Specifically, this work shows that RGB images carry most of the information relevant to the task of general change detection and that most significant improvements can be obtained by better exploiting these RGB bands. Additional multispectral bands can still be used in specific contexts, for instance in the form of known spectral indices but otherwise complicate the learning task by adding more noise than useful information which creates a need for more complex models and larger datasets. As to the unsupervised and semi-supervised methods, this work shows that they can compare to the synthetic dataset-trained approach while not outperforming it. Nevertheless, our appproach can provide good pretrained weights to be tuned for a more specific task of change detection.



Access Thesis_Pirenne_Thomas.pdf
Size: 61.57 MB
Format: Adobe PDF


Access Abstract.pdf
Size: 134.53 kB
Format: Adobe PDF
Access Pipeline_ConvCN.pdf
Description: Training flowchart for the ConvCN model.
Size: 86.25 kB
Format: Adobe PDF
Access Pipeline_ConvGDCN.pdf
Description: Training flowchart for the ConvGDCN model.
Size: 101.67 kB
Format: Adobe PDF
Access Pipeline_GDCN.pdf
Description: Training flowchart for the GDCN model.
Size: 118.31 kB
Format: Adobe PDF


  • Pirenne, Thomas ULiège Université de Liège > Master ingé. civ. info., à fin.


Committee's member(s)

  • Van Droogenbroeck, Marc ULiège Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Télécommunications
    ORBi View his publications on ORBi
  • Wehenkel, Louis ULiège Université de Liège - ULiège > Dép. d'électric., électron. et informat. (Inst.Montefiore) > Méthodes stochastiques
    ORBi View his publications on ORBi
  • Total number of views 36
  • Total number of downloads 0

All documents available on MatheO are protected by copyright and subject to the usual rules for fair use.
The University of Liège does not guarantee the scientific quality of these students' works or the accuracy of all the information they contain.