Detection of islands and droplets on smectic films using machine learning

Abstract

Machine learning techniques have been developed to identify inclusions on the surface of freely suspended smectic liquid crystal films imaged by reflected light microscopy. The experimental images are preprocessed using Canny edge detection and then passed to a radial kernel support vector machine (SVM) trained to recognize circular islands and droplets. The SVM is able to identify these objects of interest with an accuracy that far exceeds that of conventional tracking software, especially when the background image is non-uniform or when the target features are in close proximity to one another. This method could be applied to tracking objects in a variety of visually inhomogeneous biological and soft matter environments, in order to study growth dynamics, the development of spatial order, and hydrodynamic behavior.

Machine learning methods for liquid crystal research: phases, textures, defects and physical properties

“…A recently developed radial support vector machine has shown high performance in detecting isotropic droplets and pancake-like islands, as well as their combinations. 88 SVM proved to be particularly useful in the case of heterogeneous smectic A films, where the efficiency of the widely used TrackPy software for object tracking did not exceed 25% in detecting droplets in close contact and was less than 10% in the case of nested objects (droplets within islands), while SVM showed an efficiency of 100% and about 90%, respectively, when analyzing the same images (Fig. 5(b)).…”

Possibilities and limitations of convolutional neural network machine learning architectures in the characterisation of achiral orthogonal smectic liquid crystals

“…This work is to a large extent connected to the identification of topological defects in experimental 28 and simulated nematic textures, 29 thus related to object recognition. 30 Closely relating to this is an investigation of machine learning detection of bubbles and islands in free-standing smectic films, 31 and work on active nematics relating to hydrodynamics. 32…”

“…This work is to a large extent connected to the identification of topological defects in experimental 28 and simulated nematic textures, 29 thus related to object recognition. 30 Closely relating to this is an investigation of machine learning detection of bubbles and islands in free-standing smectic films, 31 and work on active nematics relating to hydrodynamics. 32 Further machine learning studies were connected to theoretical predictions of the molecular ordering of binary mixtures of molecules with different length, 33 the self-assembled nanostructures of lyotropic liquid crystals, 34 and the local structure of liquid crystalline polymers.…”

Prediction of the Structural Color of Liquid Crystals via Machine Learning

“…In other applications, images of shear-sensitive coatings have been analyzed with machine learning to determine the magnitude and direction of the shear stress based on the observed color [11]. Machine learning has also been used for detection of oil droplets in inhomogeneous smectic liquid crystal films based on images [26]. Scattering patterns of liquid crystal droplets in flow cytometry have also been analyzed using a machine learning approach to identify lipopolysaccharides from three bacterial organisms and predict their concentration in aqueous media.…”