Areal imaging and deep learning for forest assessment

27 Janvier 2023

---

Catégorie : Post-doctorant

---

Post-doc position

Title: Aereal imaging and deep learning for forest trees assessment

Keywords: Multispectral imaging, remote sensing, image processing, deep learning, data fusion, forest, environment.

 

Context: Forest is one of the richest environments in terms of biodiversity. It constitutes a reservoir and a shelter for fauna and wild flora. Forest helps to stabilize the local and general climate by acting on humidity, temperature and wind. It allows to fight against erosion, avalanches, floods, groundwater pollution, it improves quality of underground water, reduces energy costs for water purification. However, forest is a fragile environment that is particularly subject to hazards of climate change. It is therefore necessary to better understand forest and to act to promote its development and its adaptation to climate change through precision silviculture.

Objective: the work consists in the development of software tools to recognize and measure, from satellite and drone photos, characteristics of trees, growth of trees, impacts of climate change and alert on risks of dieback. The goal is to propose new methods based on deep learning approaches for multimodal data fusion, namely 3D and multispectral information. Since we face a scarcity of labeled data and a large amount of unlabeled data, we will focus on approaches that combine self-supervised learning with supervised and semi-supervised learning.

 

Profile: PhD degree with experience in machine learning for analysis of images and data with different modalities.

Duration: ~18 months

Location: PRISME Laboratory, INSA CVL, 88 Boulevard Lahitolle, 18022 Bourges.

Applications: CV to be sent to: adel.hafiane@insa-cvl.fr ; raphael.canals@univ-orleans.fr