Towards Vision-based Markerless Tracking and Force Analysis of Weightlifting-specific Technical Movements

Weightlifting, as a pivotal sport for our nation in the pursuit of Olympic gold medals, plays a crucial role in elevating our competitive sports standards. However, as a high-intensity sport, it demands exceptional technical precision from athletes. Traditional training methods often rely heavily on the subjective experience of coaches and athletes’ self-assessment, resulting in the problems such
as high variability, delayed feedback, and inadequate detail capture. These challenges are exacerbated in local training teams where coaching resources are limited, posing substantial barriers to the
identification and cultivation of emerging talents.

The existing computer-assisted training systems, such as marker-based motion capture systems, are quite expensive, thus hindering the widespread applicability. Furthermore, such a system fails to
provide comprehensive force analysis for effective training. Therefore, it is desired to develop an innovative approach to enhancing the precision and immediacy of technical movement analysis in weightlifting.

This project aims to harness cutting-edge computer vision technology to develop a markerless system for tracking weightlifting techniques and analyzing biomechanical forces. By accurately capturing critical postures throughout the movement and evaluating the forces acting on various body segments, this research will establish a scientific foundation for weightlifting-specific training. Besides that, the underlying techniques developed through this proposed project are also applicable to the other competitive sports disciplines.