Research progress on model and analysis method of jumping posture in figure skating

Abstract

This paper reviews the human skeletal dynamics model employed in figure skating pose analysis. The prevailing contemporary technological framework is predicated on the integration of artificial observation assistance, miniature wearable unit data, and dynamic human pose recognition, a suite of technical methodologies. This paper assesses the technical tools and advantages and disadvantages of these three approaches, with a focus on neural network-based multi-fusion 3D pose recognition. Finally, the study offers a discussion of the potential for hardware and software performance improvements to facilitate more accurate analysis of various figure skating models. This, in turn, could guide and train athletes to achieve more difficult movement breakthroughs. Moreover, such improvements could enhance the enjoyment of sports competitions.