Mobile real-time monitoring system based on human action recognition

Abstract

Mobile real-time monitoring system which is based on the wireless network meets the requirements of users who want to monitor and manage the home in a mobile scene or an emergency scene. This chapter designs and implements a real-time monitoring system based on mobile terminal. At the same time, human action recognition is applied to the system. A method combining the 3D skeleton shape histogram with dynamic time warping (DTW) is proposed to improve the accuracy of recognition.