Smart Cushion Based on Pressure Sensor Array for Human Sitting Posture Recognition

Abstract

A smart cushion based on a pressure sensor array is designed and manufactured to recognize human sitting posture. The smart cushion collects pressure images for convolutional neural network training to recognize 5 human sitting postures. In a series of experiments, a total of 16,300 pressure images were collected by 5 subjects to perform recognition tasks. By collecting pressure images multiple times to optimize the time independence of the data set, and the accuracy achieved 0.978. While ensuring comfortness, this smart cushion takes into account both the high accuracy of sitting posture recognition and the development prospects of visualization applications.

Liangqi Yuan

PhD Student