[Defended thesis] Y. Jean Eudes Gbodjo: Leverage Multi-Source Remote Sensing data via machine learning to improve Crop Monitoring Systems

Leverage Multi-Source Remote Sensing data via machine learning to improve Crop Monitoring Systems

I'm Yawogan Jean Eudes Gbodjo, an Irstea doctoral student at the UMR Tetis (Territoires, Environnement, Télédétection et Information spatiale) in Montpellier.

I'm a graduate of the Master 2 ScIences Géomatique en environneMent et Aménagement (SIGMA) at the Université de Toulouse Jean Jaurès. I'm a geomatician, specializing in Remote Sensing and Geographic Information Systems, with a strong interest in Computer Science. This is one of the reasons why I decided to do this thesis.

I'm working on new machine/deep learning methods that can take advantage of the wealth of information available today, thanks in part to earth observation satellites, to improve crop monitoring systems. In particular, we want to better characterize cultivated areas (crop types, surface areas) and their yields. This work is in line with a major current issue in agriculture: how can we guarantee food security for our ever-growing populations, while promoting sustainable agriculture that preserves our ecosystems and biodiversity in the face of the environmental impacts of climate change, which are already making themselves felt?

• Starting date: November 2018
• University: University of Montpellier
• PhD school: I2S Information Structures Systèmes
• Scientific fiels: IT
• Thesis management: Dino Ienco, Inrae, UMR Tetis
• Thesis supervisors:  Dino Ienco, Inrae, UMR Tetis & Louise Leroux, Cirad, UMR Aida
• Funding: #DigitAg – Inrae
• #DigitAg: Cofunded PhD – Axe 5 – Challenges 6 & 8

Keywords: Agricultural production, Land use, Yield estimation and forecasting, Multi-source, multi-temporal and multi-scale data, Radar and optical images, Supervised learning, Recurrent neural networks, Convolutional neural networks

Abstract: Face to population increasing and the environmental impacts of climate change, ensuring the food security of populations while promoting sustainable agriculture that preserves terrestrial ecosystems and biodiversity (Goals 2 and 15 of the UN Sustainable Development) becomes a major challenge for the future of our society. As satellite missions multiply (eg Sentinel), various sources of information are now available to better characterize agricultural systems and associated practices at the regional, national and global scales. At the same time, the integration of these various sources of data, especially for agronomic issues, remains a real challenge. The aim of this Ph.D. is to propose new machine learning approaches particularly with deep learning to integrate the multi-source and the multi-temporal information provided by optical and radar time series and very high spatial resolution imagery in order to improve the characterization of cultivated areas and the estimation of agricultural yields from data collected in the field. The developed approaches will be evaluated with a cross look at contrasting sites in terms of crop systems and / or agricultural practices (France, Senegal).

Jury compound:

• Bertrand LE SAUX,    Rapporteur, Chargé de recherche, HDR  lab ESA
• Clément MALLET, Rapporteur,  Cadre scientifique des EPIC, HDR IGN, LASTIG Université Gustave Eiffel Paris France
• Thomas CORPETTI,  Examinateur, Directeur de recherche, LETG Université Rennes 2
• Germain FORESTIER, Examinateur,  Professeur, IRIMAS, Université Haute-Alsace
• Mme Carmen GERVET, Examinateur, Professeur, Espace-Dev, Université de Montpellier
• Mme Laurence HUBERT-MOY,  Examinateur, Professeur, LETG Université Rennes
• Dino IENCO, Directeur de these , Chargé de recherche, HDR, Université de Montpellier
• Mme Louise LEROUX, Co-encadrante, Chargé de recherche, PhD , UPR AIDA CIRAD

Download the thesis manuscript

Papers in international journals

Censi A.M., Ienco D., Gbodjo Y.J.E., Pensa R.G., Interdonato R., Gaetano R. (2021) Attentive Spatial Temporal Graph CNN for Land Cover Mapping from Multi Temporal Remote Sensing Data, IEEE Access

Ienco D., Eudes Gbodjo Y.J., Gaetano R., Interdonato R. (2020) Weakly supervised learning for land cover mapping of satellite image time series via attention-based CNN, IEEE Access

Ienco D., Gbodjo Y.J.E., Gaetano R., Interdonato R. (2020) Generalized Knowledge Distillation for Multi-Sensor Remote Sensing Classification: An Application to Land Cover Mapping, ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences

Gbodjo, Y.J.E.; Ienco, D.; Leroux, L.; Interdonato, R.; Gaetano, R.; Ndao, B., 2020, Object-Based Multi-Temporal and Multi-Source Land Cover Mapping Leveraging Hierarchical Class Relationships

Yawogan Jean Eudes Gbodjo; Dino Ienco; Louise Leroux, 2019, Toward Spatio–Spectral Analysis of Sentinel-2 Time Series Data for Land Cover Mapping

Eudes Gbodjo Y.J., Leroux L., Gaetano R., Ndao B., 2019, RNN-based multi-source land cover mapping: An application to West African landscape

Contact :   jean-eudes.gbodjo [AT] inrae.fr​ / Tel:  04.67.54.87.54

Social networks:   GitHub – LinkedIn – Twitter