ABSTRACT
Hand hygiene has become even more im-portant in light of the COVID-19 pandemic, where hands are one of the high-risk transmission routes. Existing hand-hygiene education is focused on one-time training and does not ensure that correct handwashing procedures are undertaken. Our study, therefore, proposes a hand-hygiene education and facilitation system. Compared to previous systems, through an external RGB camera with our proposed image preprocessing and use the 3-D convo-lution and convolutional long short-term memory (Con-vLSTM) models to detect correctness of handwashing postures, which also facilitates children's ability to wash their hands properly through an on-screen tutorial. It also encourages children to develop good handwashing habits through a positive competition and reward system, and helps teachers to understand children's learning pro-gresses. The experimental results showed that the model was able to identify handwashing postures in real-time with 95.12% accuracy in a realistic and variable environ-ment. © 2022 IEEE.