Research and Implementation of an AI-Based Human Posture Recognition System
DOI:
https://doi.org/10.5281/zenodo.15187784Keywords:
Artificial Intelligence, Human Posture Recognition, Multi-Source Data Fusion, Deep Learning; Sports TrainingAbstract
This study aims to construct a human posture recognition system based on artificial intelligence, with the goal of enhancing the accuracy and real-time performance of posture recognition. The system adopts a modular design that encompasses data collection, preprocessing, feature extraction, posture recognition, and result evaluation. By integrating multi-source data through a fusion strategy that combines visual and inertial sensor data, the system utilizes deep learning algorithms to achieve precise posture recognition. Experimental results demonstrate that the system excels in both sports training and medical rehabilitation fields. It can provide personalized training recommendations for athletes and assist coaches in optimizing training programs. Additionally, it helps doctors and therapists accurately monitor patient recovery progress, thereby improving rehabilitation outcomes. In the future, continuous optimization of the system will be pursued, expanding its application domains and providing stronger technical support for the intelligent development of related industries.
References
Rui Liang, Xing Luo, Bowen Zhao, Jiangxin Yang & Yanpeng Cao. (2025). Enhancing active steady-state non-line-of-sight recognition via a guided generative adversarial framework. Optics and Laser Technology, 183, 112269-112269.
Sujoy Karmakar, Subhadeep Das, Subir Podder, Hriday Kumar Basak, Soumen Saha & Abhik Chatterjee. (2024). Synthesis, characterization and solvent effects on the electronic properties and non-linear optical activity of acridine-1, 8-diones-proline based fluorescent probe for visual recognition and selective detection of OH− ion. Journal of Molecular Liquids, 401, 124661-.
K. Aarthy & A. Alice Nithya. (2024). Automated yoga pose recognition using enhanced chicken swarm optimization with deep learning. Multimedia Tools and Applications, 83(38), 1-23.
Amit Baghel, Alok Kumar Singh Kushwaha & Roshan Singh. (2025). Automated Human Action Recognition with Improved Graph Convolutional Network-based Pose Estimation. International Journal of Pattern Recognition and Artificial Intelligence, (prepublish),
Youssef Nkizi & Ornwipa Thamsuwan. (2024). Fall Risk Assessment in Active Elderly Through the Use of Inertial Measurement Units: Determining the Right Postural Balance Variables and Sensor Locations. Applied Sciences, 14(23), 11312-11312.
Xiongfeng Li, Limin Zou & Haojie Li. (2024). Tai Chi Movement Recognition and Precise Intervention for the Elderly Based on Inertial Measurement Units and Temporal Convolutional Neural Networks. Sensors, 24(13), 4208-4208.
Wei Zhao, Lei Wang, Yuanzhe Li, Xin Liu, Yiwen Zhang, Bingchen Yan & Hanze Li. (2024). A Multi-Scale and Multi-Stage Human Pose Recognition Method Based on Convolutional Neural Networks for Non-Wearable Ergonomic Evaluation. Processes, 12(11), 2419-2419.
Yang Min, Zhang Yunong, Zhou Xuefeng & Hu Haifeng. (2021). Pose control of constrained redundant arm using recurrent neural networks and one-iteration computing algorithm. Applied Soft Computing Journal, 113(PB),
Hong Zhao, Bojing Du, Yongju Jia & Hui Zhao. (2025). DanceFormer: Hybrid transformer model for real-time dance pose estimation and feedback. Alexandria Engineering Journal, 121, 66-76.
Haoran Zhan, Jiange Kou, Yuanchao Cao, Qing Guo, Jiyu Zhang & Yan Shi. (2024). Human Gait phases recognition based on multi-source data fusion and BILSTM attention neural network. Measurement, 238, 115396-115396.
Mohammad Mahdi Rusta, Seyyed Arash Haghpanah, Sajjad Taghvaei & Ramin Vatankhah. (2024). Adaptive neuro-fuzzy sliding mode control of the human upper limb during manual wheelchair propulsion: estimation of continuous joint movements using synergy-based extended Kalman filter. Neural Computing and Applications, 36(28), 17375-17416.
Jiaying Lan, Guoheng Huang & Lianglun Cheng. (2019). Fast 3D Post Estimation of Human Based on Optical Flow and Particle Filter. Journal of Software, 437-448.
Jiale Qiao, Jindong Zhang, Yuze Wang, Chenhui Yu & Sai Gao. (2019). A Multi-sensor Fusion System for Embedded Devices Considering by Sensor Reliability.(eds.)Proceedings of 2019 2nd International Conference on Mechanical Engineering, Industrial Materials and Industrial Electronics(MEIMIE 2019) (pp.88-91). Jilin University;
Kazuki HAYASHI & Daisuke TANAKA. (2023). Effectiveness of Feature Extraction System for Multimodal Sensor Information Based on VRAE and Its Application to Object Recognition: Special Section on Deep Learning Technologies: Architecture, Optimization, Techniques, and Applications. IEICE Transactions on Information and Systems, E106.D(5), 833-835.
Kaige Ding, Zhinan Zhao, Siyuan Ma, Yanqing Qiu, Tingting Lang & Ting Chen. (2024). Accelerating optimization of terahertz metasurface design using principal component analysis in conjunction with deep learning networks. Physical Communication, 66, 102452-102452.
Sheng Bo, Chen Linfeng, Cheng Jian, Zhang Yanxin, Hua Zikai & Tao Jing. (2024). A markless 3D human motion data acquisition method based on the binocular stereo vision and lightweight open pose algorithm. Measurement, 225, 113908.
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Feng Miao, Qiang Zhang, Jun Yan
This work is licensed under a Creative Commons Attribution 4.0 International License.
