Back to Search
Start Over
Design and Optimization of Motion Training System Assisted by Human Posture Estimation Algorithm.
- Source :
- Scientific Programming; 5/25/2022, p1-12, 12p
- Publication Year :
- 2022
-
Abstract
- With the rapid development of computer technology and electronic information technology, the sports training system no longer depends on the traditional algorithm for operation support, and various advanced posture algorithms are emerging. At the same time, it also further optimizes the intelligence and accuracy of the sports training algorithm. As an advanced algorithm combined with virtual reality technology, human posture estimation algorithm plays an obvious role in optimizing the effect of sports training. This paper will design a motion training system based on the optimized and improved human posture trajectory algorithm, use the depth image correlation theory to solve the problem of non-Gaussian noise crosstalk in the depth image of the traditional human posture algorithm in principle, improve the accurate feature extraction of the depth image by the algorithm, and solve the problem of human feature redundancy, so as to further improve the accuracy of the establishment of a single human model; on the problem of multi-person posture estimation algorithm, this paper proposes a high-resolution multi-person posture high-precision network model and adds the focus mechanism. Based on this, this paper realizes the high-precision and high-speed modeling of multi-person posture, so as to provide an accurate model for the multi-person function of sports training system and improve the efficiency of the algorithm. In the experimental part, this paper takes tennis as a typical case to design the sports training system and experiments based on the system designed in this paper. The experimental results show that the system under the proposed algorithm has obvious advantages in accuracy and training effect. [ABSTRACT FROM AUTHOR]
Details
- Language :
- English
- ISSN :
- 10589244
- Database :
- Complementary Index
- Journal :
- Scientific Programming
- Publication Type :
- Academic Journal
- Accession number :
- 157072171
- Full Text :
- https://doi.org/10.1155/2022/6839768