Wearable device in college track and field training application and motion image sensor recognition

Abstract

With the continuous advancement of microelectronics and sensor technology and the development of deep learning theories, the knowledge of human behavior based on wearable devices has become a new research direction that reflects the nature of human movement more than that of human behavior. Computer vision human action recognition based on wearable devices is not affected by specific scenes and limited time, has low energy consumption and low cost, and is suitable for promotion. From the perspective of functional requirements, user actions, technical layouts of major companies, policy directions, and other aspects of wearable devices, sensor-related technologies are factors that limit the development of the wearable industry, and the development of sensor technology stimulates the longest development of wearables Equipment industry. This article combines human posture measurement with the basic knowledge of the human body model to construct a parametric and changeable human posture model. According to the basic theory of wearable human posture monitoring technology, use quota Niang and Euler angles to determine the descriptive relationship between the collected data and the human posture. Based on the existing theoretical foundation and the university s land training practice, the overall design of the posture data collection module, wireless communication module and system including application software is proposed. Finally, it is analyzed according to the joint analysis model of the human body. Experimental results show that this system can achieve the purpose of providing training items for track and field athletes.