Transfer Learning of Photometric Phenotypes in Agriculture Using Metadata
This addresses the challenge of accurate phenotype estimation for agriculture, enabling better decisions on yield quality and breeding, but it is incremental as it builds on existing methods with metadata integration.
The paper tackles the problem of estimating photometric plant phenotypes from images under varying field conditions by embedding metadata about capturing conditions into a network, resulting in improved accuracy for tomato hue and chroma compared to a state-of-the-art deep CNN and a human expert.
Estimation of photometric plant phenotypes (e.g., hue, shine, chroma) in field conditions is important for decisions on the expected yield quality, fruit ripeness, and need for further breeding. Estimating these from images is difficult due to large variances in lighting conditions, shadows, and sensor properties. We combine the image and metadata regarding capturing conditions embedded into a network, enabling more accurate estimation and transfer between different conditions. Compared to a state-of-the-art deep CNN and a human expert, metadata embedding improves the estimation of the tomato's hue and chroma.