CNN-based visual analysis to study local boiling characteristics

Abstract Visual analysis allows an estimate of different local boiling characteristics including bubble growth rate, departure diameters and frequencies nucleation, nucleation site density evolution bubbles dry spots in time. At the same time, visual determination presented case big amounts data requires development appropriate software which will allow not only location, but also their sizes based on high-speed video. The problem can be solved by using instance segmentation approach a convolutional neural network. In work Mask R-CNN network architecture was used for estimation characteristics.