Jednoznačná identifikace osob na senzorické podlaze

Abstract

This thesis deals with the design and implementation of a system for identifying individuals in the environment of the residential laboratory HEALTH.Lab through the analysis of raw data obtained from the SensFloor sensor floor. The goal is to create a unique identifier for individuals or groups of people based on their interaction with the floor, with the main monitored parameter being the capacitive response caused by the type of footwear, the user's weight, and walking style. As part of the solution, a methodology for data collection and processing was designed, experimental measurements were carried out in the laboratory's real environment, a software tool was created for data visualization and classification, and various classification algorithms were tested. The best results were achieved using the XGBoost model, which demonstrated high accuracy in distinguishing between selected groups. The contribution of this work is the possibility of indirect, non-invasive identification of individuals without the need for cameras or wearable devices, with potential applications in assistive technologies, smart homes, and healthcare. The results confirm that the system is capable of not only detecting presence but also distinguishing individuals or their categories based on the physical properties of the interaction with the floor.