REAL-TIME YOGA ASSISTANCE SYSTEM USING COMPUTER VISION TECHNIQUE

Abstract

The Real-Time Yoga Assistance System using Computer Vision Technique is an intelligent and interactive platform designed to support yoga practitioners by providing real-time posture detection and corrective feedback. Many individuals practice yoga without expert supervision, which often results in incorrect posture alignment and increased risk of injury. This system addresses these challenges by utilizing computer vision and machine learning algorithms to analyze human body movements through live video streams. Key body landmarks are detected and compared with predefined standard yoga postures to evaluate accuracy. Technologies such as Python, OpenCV, MediaPipe, and Flask are employed for implementation. The system enhances safety, improves posture accuracy, and makes yoga practice accessible to users of all skill levels.