Intelligent Monitoring Technology and Safety Protection Analysis of Elderly Care Robots Based on Multi-Modal Perception
Abstract
. As the process of aging the population of the planet becomes more and more significant, more of the elderly people will live alone, and elderly care robots which will assist them 24/7 can be used. This paper aims at exploring and discussing the development and use of essential technologies in hardware and software in three ways: in terms of vision, hearing, and vital signs monitoring. Interferometric fiber optic microphones and remote photoelectric plethysmographs can be incorporated on the hardware side and provide new solutions to an auditory reception and contactless measurement of elderly care robots, and several visual hardware accelerators will significantly enhance visual recognition capabilities. Deep neural networks, Transformer tracking and prediction models, which use local multi-scale features (Lo TPM) and the Kanade-Lucas-Tomasi (KLT) tracking algorithm also use the hardware, on the software side. This paper will attempt to offer a source to upcoming researchers.