MH 09-MoCap in Sports

Quantifying sports performance with markerless motion tracking

MH - Sport and Exercise Science

Chair of Human Movement Science

human movement science

Computer Science/ InformaticsHealth SciencesSport Sciences

biomechanics/human movement science/kinesiology/biopsychology

Planned project location:

CiO - Campus at the Olympic Park

Dr

Philipp

Gulde

philipp.gulde@tum.de

+49.89.289.24550

Performance analysis in sports still heavily relies on subjective judging by an expert trainer. Although, kinematic analyses have the advantage of being quantitative, sensitive, and objective, the use is limited to the highest level of competitive sports and is considered cumbersome. Recent advances in machine learning based markerless motion tracking offer the opportunity to capture meaningful body kinematics in sporting contexts at little cost and effort. In this project, we aim to develop a user-interface that supports athletes with automated kinematic analyses of movements, considering outcomes like maximum velocities (e.g., in boxing) and distances (e.g., in jumping), as well as more abstract ones that quantify the quality of a movement (e.g., jumping technique in basketball players or figure configurations in gymnastics).
Your tasks will be:
1. Explore the literature on kinematic analyses, in sports and other contexts (e.g., clinical)
2. Develop a user-interface for automated kinematic analyses within a defined sporting scenario.
3. Validate your program (and its kinematic features) with a gold-standard optoelectronic motion capturing system and based on expert ratings.

30

3 years of bachelor studies completed

Basics in:
1. human movement science
2. kinematic analyses
3. coding/programming
Interest in:
1. sports
2. performance analysis

Basics in:
1. empirical research (incl. statistics)
2. literature research