Ambulatory Human Posture Detection using MARG Sensor Arrays and Madgwick Filter

Abstract

This report proposes a novel and low cost method of calculating and collecting real time human posture and orientation data with the help of MARG (Magnetic, Angular Rate and Gravity) sensors, micro controllers, Madgwick filter and an on board mobile computer. With the help of MARG sensor arrays, the proposed method overcomes the movement and environmental restrictions of traditional vision based posture data collection systems. Madgwick filter, being very light weight but accurate, allows the development of the whole system at a cost cheaper than other similar sensor based alternatives, using commercially available simple hardware. The low cost implementation, accuracy and mobility provided by the system can open new opportunities for the use of such systems in research, medical, input modality, sports, military, computer graphics and other fields, specially in places where cost and mass production are crucial factors.