2017/09/30 by Shubhra Aich, Anique Josuttes, Aich, Shubhra +15
Agricultural and Biological Sciences · Chemistry · Environmental Science · #Computer Vision and Pattern Recognition (cs.CV) #FOS: Computer and information sciences #Remote Sensing in Agriculture #Smart Agriculture and AI #Spectroscopy and Chemometric Analyses
paper · pdf · doi:10.48550/arxiv.1710.00241
openalex publication_date 2017/09/30 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
In this paper, we investigate estimating emergence and biomass traits from\ncolor images and elevation maps of wheat field plots. We employ a\nstate-of-the-art deconvolutional network for segmentation and convolutional\narchitectures, with residual and Inception-like layers, to estimate traits via\nhigh dimensional nonlinear regression. Evaluation was performed on two\ndifferent species of wheat, grown in field plots for an experimental plant\nbreeding study. Our framework achieves satisfactory performance with mean and\nstandard deviation of absolute difference of 1.05 and 1.40 counts for emergence\nand 1.45 and 2.05 for biomass estimation. Our results for counting wheat plants\nfrom field images are better than the accuracy reported for the similar, but\narguably less difficult, task of counting leaves from indoor images of rosette\nplants. Our results for biomass estimation, even with a very small dataset,\nimprove upon all previously proposed approaches in the literature.\n