Abstract
In recent years, contactless monitoring has become the most popular research topic in healthcare monitoring systems. By monitoring vital signs continuously, we will know about our health conditions. We can identify many diseases based on our vital signs. Currently, there are RFID-based, radar-based, Wi-Fi-based, and camera-based techniques to monitor our vital signs. However, these techniques might not provide robust results under certain conditions. For example, camera-based breathing sensing may not work robustly if the room has low brightness. Wi-Fi-based breathing sensing is not robust when there are disturbances in the environment. So, we proposed a model to improve the robustness of vital sign monitoring by combining different sensing techniques.
In this project, we develop a system to monitor our breathing using Wi-Fi CSI data and a camera. Later, we use a neural network model which combines Wi-Fi-based breathing sensing and camera-based breathing sensing to improve the robustness.
This research mainly focuses on extracting breathing data from Wi-Fi CSI. We also use two Raspberry Pi units and a router to obtain the CSI data. We also use a low-resolution camera to capture breathing information through video. For the code, we use Python libraries, and MATLAB.