This thesis deals with a cutting-edge research topic in information science: the possibility of using random patterns formed by tiny spheres of liquid crystals integrated into common materials, as an innovative solution for physical object identification. The thesis aims to generate and analyze images of these liquid crystals, which, due to their optical properties, generate random patterns, and test model-based and data-driven approaches for recognizing the said crystals. The activities of the thesis have been conducted in the labs of the Center for Security, Reliability, and Trust (SnT) (https://wwwen.uni.lu/snt) of the University of Luxembourg, under the co-supervision of Dr. Gabriele Lenzini.