Fig. 7
From: Machine learning based multi-parameter droplet optimisation model study
Droplet pixels segmented out of the inkjet device as well as the microfluidic device droplet image watershed are quantised by an objective function \(\:l\left(x\right)\). The objective function consists of a geometric loss \(\:{L}_{geom}\), a yield loss \(\:{L}_{yield}\), and a size uniformity loss \(\:{L}_{size}\), based on which all droplet pixel objective function values are calculated for each image.
