Estimating physical properties from liquid crystal textures via machine learning and complexity-entropy methods

Abstract

Imaging techniques are essential tools for inquiring a number of properties from different materials. Liquid crystals are often investigated via optical and image processing methods. In spite of that, considerably less attention has been paid to the problem of extracting physical properties of liquid crystals directly from textures images of these materials. Here we present an approach that combines two physics-inspired image quantifiers (permutation entropy and statistical complexity) with machine learning techniques for extracting physical properties of nematic and cholesteric liquid crystals directly from their textures images. We demonstrate the usefulness and accuracy of our approach in a series of applications involving simulated and experimental textures, in which physical properties of these materials (namely: average order parameter, sample temperature, and cholesteric pitch length) are predicted with significant precision. Finally, we believe our approach can be useful in more complex liquid crystal experiments as well as for probing physical properties of other materials that are investigated via imaging techniques.

Areas of opportunity related to design of chemical and biological sensors based on liquid crystals

“…AlexNet was trained using millions of images found on the Internet, thus mimicking how humans categorise new objects based on prior information. In the context of research involving LCs, AlexNet-based frameworks have been used to identify LC phases and predict the order parameters of simulated nematic LCs [34,35]. CNNs have also been used to identify the pitch length of simulated samples of cholesteric LCs [34].…”

“…In the context of research involving LCs, AlexNet-based frameworks have been used to identify LC phases and predict the order parameters of simulated nematic LCs [34,35]. CNNs have also been used to identify the pitch length of simulated samples of cholesteric LCs [34].…”

Collective dynamics of stock market efficiency

“…By definition, the normalized permutation entropy is bounded to the interval 0 ≤ H ≤ 1 , where values close to the upper bound ( H ≈ 1 ) occur for random time series and lower values of H indicate that the time series exhibits a more complex ordering dynamics. Mainly because of its simplicity, discrimination capabilities, and fast computational evaluation, the permutation entropy framework has successfully been used in many applications [44][45][46][47][48][49][50][51][52] .…”

Determining liquid crystal properties with ordinal networks and machine learning

“…This is because liquid crystals are birefringent, so that polarized optical microscope imaging often suffices to determine many different properties of these materials [13,14]. Despite that, the use of machine learning methods in liquid crystals remains surprisingly limited, and indeed only a few works have tried to directly associate physical properties of these materials with their optical textures [15,16]. As in other machine learning problems (regressions or classifications) involving images, one can learn the underlying physics of liquid crystals by extracting features from optical textures and training algorithms with a set of examples consisting of images and their associated physical properties.…”