Застосування технологій комп'ютерного зору для діагностики та реабілітації рухових розладів верхніх кінцівок

Abstract

This article covers the development and application of markerless limb key point detection systems based on artificial intelligence technologies, with a special focus on biomedical engineering for detecting upper limb movement disorders. The main attention is focused on the problems of speed and efficiency of real time detection systems when using on-device models on a mobile device. As an alternative, MediaPipe Hands, developed by Google, is proposed, which demonstrates the ability to identify key points of the hand in real time, indicating the wide potential of MediaPipe for integrating various artificial intelligence algorithms in mobile applications for diagnosing and analyzing motor skills.