Particle Shape Recognition with Interferometric Particle Imaging Using a Convolutional Neural Network in Polar Coordinates

A convolutional neural network (CNN) was used to identify the morphology of rough particles from their interferometric images. The tested had shapes sticks, crosses, and dendrites as well Y-like, L-like, T-like shapes. conversion images polar coordinates enabled particle shape recognition despite random orientations sizes particles. For non-centrosymmetric (Y, L, T), CNN not disturbed by twin image problem, which would affect some classical reconstructions based on phase retrieval algorithms. 100% rate obtained.