Real-time detection of aerobics posture based on strain sensor

Abstract

AbstractThe Internet of Things is a network that realizes the intelligent connection between things. It is another major industry after the computer industry and the Internet industry. It is also a collection of computer technology, communication technology, sensor technology, storage technology, and many other leading technologies in an integrated industry, the application industry is very wide. This article mainly introduces the real-time detection of aerobics posture based on strain sensors. This paper proposes a real-time detection method of motion posture based on strain sensors, using flexible rods to simulate the joints of aerobics athletes. Through the survey of the motion posture by the strain sensor, the relevant data and parameters are collected, and then the data are processed, and the data are finally transmitted to the terminal. Since the stress of a specific point on the surface of the flexible rod is proportional to the curve of that point, strain gauges can be used to detect curve changes at multiple points in real time and draw the change curve on the computer. The experimental structure of this paper shows that the use of BP neural network to process the measurement data improves the accuracy of real-time detection. The error distance between the sampling point and the actual value is less than 0.3 cm. In addition, through the real-time detection of the motion posture, it can be concluded that good Body posture is the basic requirement of aerobics, and it is of great significance to improve the effect of improving the quality of movement and artistic expression.