Main subject area: Phenotyping; deep learning, UAV-based remote sensing
UAV-based remote sensing is a powerful tool for field phenotyping that enables non-destructive measurement of crop physiological traits. High spatial and temporal crop sensing data will facilitate understanding of crop physiology, which enable efficient breeding programs and better-informed decision making. UAV-based RGB or multispectral imagery will be used to train a computer vision-based deep learning model to predict crop physiological traits such as biomass, nitrogen content, and nitrogen nutrition index. This project aims not only to enhance the model prediction accuracy but also to evaluate the robustness and scalability of the trained model. The student will gain experience in methodology of data preprocessing of UAV imagery, programming in Python for implementing deep learning. The student will also have a chance to collaborate with PhD students to conduct on-going field experiment projects.
Between April and September 2024 and/or April and September 2025
Department of Agroecology, AU Flakkebjerg, Forsøgsvej 1, 4200 Slagelse
45 ECTS: Experimental theses in which the student is responsible for collection and analysis of his/her own original data
60 ECTS: Experimental theses in which the student is responsible for planning, trial design and collection and analysis of his/her own original data
Tanabe et al. (2023). Winter wheat yield prediction using convolutional neural networks and UAV-based multispectral imagery. Field Crops Research 291, 108786.
