Wearable Motion Attitude Detection and Data Analysis Based on Internet of Things

With the development of the Internet of Things technology and the continuous improvement of manufacturing processes, smart devices continue to evolve. In the existing research on human motion posture detection, there are relatively few portable wearable devices. Therefore, research on human motion posture detection based on smart wearable devices will be an important direction for future development. In this paper, a smart phone with a built-in triaxial acceleration sensor is used as a data acquisition facility to simulate the wearable device's experiment on the daily movement of the human body. The three-dimensional acceleration data during the motion is collected, and the related methods are proposed to detect the human body motion posture, and further analyze the transition between various postures, and detect the special fall posture. Starting from the two aspects of time domain and frequency domain, the analysis algorithm of human motion attitude detection is expounded. Using the time domain algorithm to detect the current motion posture of the human body, the current motion posture of the human body can be quickly detected. However, if we simply use time-domain-based detection methods, it is easy to cause errors in attitude detection. In order to improve this detection error, we combine the frequency domain analysis method, use the acceleration modulus algorithm and perform FFT (Fast Fourier Transformation) to analyze the frequency domain characteristics of the signal to distinguish the human body's motion posture and improve the basic daily posture detection of the human body. The results show that the proposed method can eliminate large noise interference and reduce the missed rate and false positive rate of human attitude detection.