Human Motion Change Detection Based On FMCW Radar

Human activity recognition is one of the technologies to monitor a person's dynamism. Generally, to introduce human activities, they still use a camera, but the camera has weaknesses that depend on the intensity of the light. Therefore, radar is highly prioritized for human activity recognition systems because it is contactless and does not depend on light. This research is an implementation of the use of FMCW Radar. This change motion detection system is made by analyzing statistical characteristics in the data and using the Change Point Detection (CPD) algorithm. With the existence of a system of identifying the moment of human movement, it is hoped that the technology for the system of recognition of human activities can be implemented in various fields. The study's results of identifying changes in motion using this FMCW radar succeeded in detecting changes in people's motion. Data testing was carried out with three main movements: standing, walking, and crawling. Using mean parameters and the CPD algorithm produces an average accuracy of 85.26%.