Showing 1–1 of 1 results for author: Karol, L

Transfer Learning of Photometric Phenotypes in Agriculture Using Metadata

Authors: Dan Halbersberg, Aharon Bar Hillel, Shon Mendelson, Daniel Koster, Lena Karol, Boaz Lerner

Abstract: 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 net… ▽ More 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. △ Less

Submitted 1 April, 2020; originally announced April 2020.

Comments: Paper presented at the ICLR 2020 Workshop on Computer Vision for Agriculture (CV4A)