Real-time accuracy control of fitness exercises

The primary objective of this study is to prevent potential musculoskeletal injuries caused by incorrectly performed fitness exercises and to ensure that exercises are executed more accurately, efficiently, and safely. In today's context, individual training has become increasingly common, while access to professional trainers remains limited. To address this need, the study proposes a real-time feedback system that analyzes exercise movements using the MediaPipe library and Pose Detection technology. Body keypoint coordinates are detected in real time, and joint angles are calculated to assess the correctness of movements. The obtained data are processed using optimization algorithms and statistical analysis techniques, enabling the evaluation of exercise form based on predefined threshold values. The developed system provides users with immediate and guiding feedback, improving exercise accuracy, reducing the dependency on personal trainers, and supporting the integration of digital technologies into the fitness domain. In this context, the study aims to contribute to the development of safer, more effective, and sustainable exercise routines for individuals.