Skip to main content
Fig. 3 | Plant Methods

Fig. 3

From: DeepFlower: a deep learning-based approach to characterize flowering patterns of cotton plants in the field

Fig. 3

Counting accuracies calculated using “plant-based counting” (top row) and “whole image-based counting” (bottom row) strategies, respectively, for individual plants on each of the 26 scanning dates (a total of 2834 data points). a and d are linear regression results between the imaging derived and manual counts. b and e are the histograms of counting errors. C and F are the histograms of relative counting errors for samples with an absolute counting error of less than 1. In c and f, the numbers on top of the bars indicate the relative improvement (over 5%) of using the “plant-based counting” strategy over the “whole image-based counting” strategy. The asterisk denotes samples that had a zero count with the manual method but a non-zero count with the imaging method

Back to article page