Data Collection and Analysis of Physical Education Teaching Practice Based on Multisensor Perception

Abstract

Because our country is constantly updating training standards, the first mechanism for assessing physical fitness is far from current needs. A good assessment process can help to stimulate student learning, identify strengths and weaknesses, and better improve sports knowledge. In this paper, based on the multisensor perception of physical education and teaching practice data collection and analysis, it is found that the original sensor data often have some defects, but the Kalman filter can be processed, which can make the data more accurate. After comparing the data, it can be found that each group of data basically has an error of 0.02–0.9. After processing, the data better reflect the changes in the measurement. With the reform of the professional evolution of boxers, higher requirements have been placed on athletes. The sensor can be continuously tested. According to the experiment, the basic probability of different sensors on the test paper can be found that the fused sensor data are 1/2, while the single sensor data are 1/6, and the data of a single sensor are much lower than the confidence of fused sensors, effectively improving the comprehensive ability of boxing.