In 2019, Wirth Lab collected microscope images and videos of paint in the drying process.
When coatings are drying, various chemical phenomena can create "convection cells," which are small, individual volumes of flowing coating. These cells are easily seen in much of the data, and the data may be analyzed to determine properties of the coating. Manual analysis can reveal direct properties such as average cell size, or cell spread, irregularity, or circularity, but perhaps computational analysis could reveal higher-order properties such as coating thickness/surface appearance, and help simplify the extraction of the simpler ones. How the convection cells settle and dry is essential in achieving certain coating properties and preventing irregularities, such as blemishes on painted autombile chassies.
The goal of this research is essentially to discover what image processing and machine learning techniques can do for the extraction of data in these images.