On-chip body posture detection for medical care applications using low-cost CMOS cameras

Abstract

A low-complexity algorithm for the computation of the geometrical moments of binary images has been developed and implemented on the limited CPU and RAM resources available in current low-cost on-chip CMOS cameras. This moment-based approach can be used to find the position and orientation of objects in the image plane, which has a wide range of applications from networked embedded robotics to the medical care field. In medical care applications it is envisioned that camera systems with on-chip processing algorithms, such as the one described in this paper, can be used to determine the posture of a human body, e.g. standing or lying down. By processing only binary images on the camera chip and then transmitting numbers related to the geometrical moments, security and privacy concerns can be taken care using simple procedures. It is expected that this type of camera systems will work in conjunction with a wireless body area network (WBAN) that measures vital signs. The final goal is to provide cost effective and timely information to the medical personnel responsible for the care of the patients.