Augmented reality technology is a cutting-edge technology of current social development, and its development promotes the development and change of people’s social production and lifestyle. In this paper, combining the key aspects of sports teaching, training, and actual combat confrontation, the construction of a VR-based immersive sports teaching environment is realized from the aspects of sports teaching materials and sports action and tactical training. Subsequently, using the bit position registration algorithm that integrates complementary filtering (CF) and untraceable Kalman filtering (UKF), as well as the TDOA localization algorithm, a VR sports-assisted training system was designed, and the performance and application effects of the system were empirically analyzed. The login page time of the system is within 3.30s, and the corresponding CPU utilization rate is 43%~50%, which is in line with the expected effect, indicating that it has good usability performance. Fusion filtering on posture angle has a lower error value than CF and UKF, which suggests higher accuracy and is suitable for posture registration in VR sports-assisted training. The TDOA localization algorithm has an error range of [0.127m, 0.389m] and a 50% error below 0.292m, which can be utilized for motion tracking in VR sports-assisted training. The experimental group using the VR sports teaching environment is significantly better than the control group using traditional teaching methods in both visual-sensory perception and technical assessment, which indicates that the VR immersive sports teaching environment constructed in this paper has practical feasibility, and provides a model reference and experience for realizing the practical application of AR technology in sports.
