Shape descriptor analysis of droplet stains on water sensitive papers

Water Sensitive Papers are among the most widely used artificial collectors worldwide for the rapid assessment of the spray distribution behavior in the field. They are primarily employed to evaluate the percentage of the covered surface area. However, at high coverage levels, the overlap between droplets introduces a significant challenge that requires further investigation. Simulations can help to better understand this phenomenon. The aims of this study were the analysis of the shape descriptors of stains resulting from single or multiple partially overlapped droplets and the estimation of the relative error in deposit measurement. Droplet stains were simulated using purposely written R functions. Four groups of images were generated, labeled from Group 1 to Group 4, each containing stains composed of one, two, three and four droplets, respectively. The simulated images were processed using ImageJ to extract shape descriptors of stains, including area, best-fitting ellipse axes, aspect ratio, Feret's diameters, circularity, roundness, and solidity. Stains from Group 1 were used as a reference. Stains from two overlapping droplets (Group 2) were used to assess how the degree of overlap affected the shape descriptors and to explore the possibility of calculating the diameters of individual droplets based on shape descriptors. Stains in Groups 3 and 4 were examined to assess the circularity index and the relative error in deposit estimation when more complex stains were involved. Within the examined stain diameter range (from 40 µm to 700 µm), all shape descriptors for reference stains deviated from their theoretical value of 1. Due to the digitization errors and randomness in stain positioning, mathematical calculation of droplet diameters for stains formed by two overlapping droplets proved to be ineffective. As stain complexity increased, circularity index decreased, and error in deposit estimation raised. These simulation results can be used to establish reference intervals for each shape descriptor and can advance the understanding of shape descriptor and droplet overlap analysis, contributing to improve pesticide application practices in terms of sustainability and safety.