Human Pose Estimation For Fitness Feedback

Bethapudi Dorka, Avvaru Ujwal, Dasari Divya; DR.Y. Chitti Babu

doi:10.62647/

Authors

Bethapudi Dorka, Avvaru Ujwal, Dasari Divya U. G Students, Dept COMPUTER SCIENCE AND ENGINEERING, St.Ann’s College Of Engineering and Technology, Nayunipalli (V), Vetapalem (M), Chirala, Bapatla Dist, Andhra Pradesh – 523187, India Author
DR.Y. Chitti Babu Associate professor, COMPUTER SCIENCE AND ENGINEERING, St.Ann’s College Of Engineering and Technology, Nayunipalli (V), Vetapalem (M), Chirala, Bapatla Dist, Andhra Pradesh – 523187, India Author

DOI:

https://doi.org/10.62647/

Keywords:

OpenCV, Media Pipe, Posture Recognition, Machine Learning. Human Tracking, Deep Learning

Abstract

This project is about creating a human pose detection system using artificial intelligence. It can recognize and show the name of a person’s body pose in real time. Many people doing exercises or yoga at home find it hard to keep the correct posture without a trainer. When the camera is turned on, the system captures live video and tracks body parts like shoulders, elbows, and knees. Using these points, it identifies which pose the person is doing and shows its name on the screen. If the pose doesn’t match any known pose, it shows “None.” The project uses Python, OpenCV, and Media Pipe to detect poses accurately. It can be helpful for fitness tracking, yoga pose detection, and movement analysis.