Tiancheng Yang soutiendra devant le jury composé de :

Abstract

To address agricultural needs, particularly in plant breeding, this thesis investigates several methods for estimating plant density, focusing primarily on wheat.

1. Using either spectral or spatial information to estimate plant density, aiming to explore their potential in terms of throughput and accuracy.
2. Using a deep learning-based plant detection method to count plants and estimate soil surface area for calculating plant density.
3. Using a deep learning-based leaf tip detection method to extract dynamic information for plant density estimation.