Grid shaped accelerometer network for surface shape recognition

In this paper a novel shape detection system is proposed consisting of accelerometer network laid out in regular grid formation. This system is specifically designed as a foundation for shape recognizing fabrics usable in medicine. Specifically, enabling production of clothing capable of posture recognition and biofeedback, for treatment of scoliosis and similar diseases. The result is a functioning 16 three-axis accelerometer prototype device and method of approximately reconstructing its shape in three dimensions with tested average error less than 1cm for surface of dimensions up to 24 × 24cm.