Recognition of Multiple Human Body Postures Based on Six-axis Sensor

In the existing methods of recognition of multiple human body postures, recognition of human body postures based on wearable sensors has recently become a research hotspot because of its advantages such as simple information acquisition, low cost, and fast transmission. Based on the monitoring data collected by the six-axis sensor, this paper performs Kalman filtering on the data, and then selects a Gradient Boosting Decision Tree model from the classification algorithms in machine learning to classify and recognize various human behavior postures. And it could be benefit to key crowds such as the elderly.