International Journal of Contemporary Research In Multidisciplinary, 2026;5(1):369-377
Ergonomic Posture Monitoring through Pose Estimation and Machine Learning
Author Name: Laishram Trinity; S Athisii Kayina; Usham Sanjota Chanu;
Abstract
Posture has a direct impact on health and daily performance, but tracking it in real time is not always easy to achieve. Traditional approaches, such as asking experts to observe, are often uncomfortable and time-consuming. With recent progress in computer vision, it is now possible to monitor posture using body landmarks. In this study, we developed a simple but effective system that uses MediaPipe to detect body landmarks and calculate joint angles from a live video feed. By analysing these angles, the system can recognise poor posture and provide immediate feedback through visual messages and auditory alerts. Tests with standard pose estimation datasets showed that the method works reliably while running efficiently on common hardware. The system can be applied in areas such as workplace ergonomics, sports practice, and rehabilitation, where continuous posture monitoring helps to reduce the risk of strain or injury.
Keywords
Posture Detection, Human Pose Estimation, MediaPipe, Real-Time Monitoring, Ergonomics, machine learning, open computer vision, random forest classifier, accuracy matrix.