Golf Swing Motion Analysis: Challenges And Solutions (P1)
The ultimate capability of a motion analysis system is determined by two major factors: motion capture capability and scope and depth of analysis allowed by the analysis software. The motion capture capability of a system depends largely on the hardware used (cameras, image processor, and controller) and the control software. With recent advancements in the real-time motion capture technologies and availability of the real-time systems, the capability of the data analysis software has become the main limiting factor.
Computation of select meaningful kinematic and kinetic parameters based on the raw position data, the outcome of motion capture, is the primary role of the analysis program. The kinds of biomechanical analyses allowed and the types of kinematic and kinetic variables provided are limited by the capability of the analysis software. For this reason, it is not rare for investigators to develop additional study-specific software programs to overcome the limitations of the general-purpose motion analysis program attached to the motion capture system. The downside of this route, however, is that each study will require a study-specific program which may not be applicable to other types of studies. When a change is made in the analysis, the program codes must be changed accordingly.
The ultimate solution to this problem is to develop a flexible, adaptive and comprehensive motion analysis program in which the scope and depth of analysis can be expanded and the need for additional study-specific programming is practically eliminated. Kwon3D (Visol, Inc., Seoul, Korea) is an example of such a program, characterized by high flexibility, adaptability, and expandability. The notion of a comprehensive analysis program is particularly attractive as the raw data now can be transferred easily from one platform to another via the file formats commonly supported by different motion captures systems, e.g. the C3D format.
A golf swing is a complex movement that presents unique challenges to the investigators, requiring advanced motion analysis methods. For instance, the outcome of a golf shot is determined by the impact conditions (the clubhead velocity and clubface orientation at impact and the impact location on the face) but computation of the impact conditions is not simple because the clubhead-ball impact does not allow placement of markers on the clubface in the dynamic trials and the sampling frequency typically used in a golf swing motion capture is not high enough to provide a sufficient time resolution. A swing analysis-specific marker set is required to locate the hand and joint centers. Definition of the local reference frames fixed to the body segments and club (head and shaft) is essential in the angular kinematics. The functional swing plane characterizes the downswing motion of a golfer but computation of its orientation and position requires a special computation algorithm. The purpose of this paper was to identify the unique challenges encountered in golf swing motion analysis, an advanced application of motion analysis, and to present solutions for these challenges.
Marker Set And Secondary Points:
A golf-specific marker set was developed with a total of 55 primary points (tracked markers) including five club markers. In addition, a total of 24 secondary points were defined and computed based on the primary points and the static trials (ball trial, club trial, and setup trial) (Table 1). Appropriate secondary point computation methods (general and joint-specific; wand, mid-point, static point, rigid body, weighted mean, etc.) were selected and configured during the body model setup process. For example, the ‘rigid body method’ (Schmidt, Disselhorst-Klug, Silny, & Rau, 1999) was used for the arm joints, while the ‘TylkowskiAndriacchi hybrid method’ (Bell, Pedersen, & Brand, 1990) was used for the hip joint in the golf-specific body model.
Data collection was conducted in the following steps:
- A static trial with the ball only was first captured (ball trial). The ball was covered with reflective tape and placed on the tee or at a pre-determined location on the mat.
- A static trial with the club only was then captured (club trial). The club markers include four shaft markers, four face markers, and one clubtoe marker (Table 1).
- A static trial with the golfer in the setup position was captured (setup trial). The clubface markers were removed in this trial.
- Dynamic (swing) trials were captured. Several body markers including the medial markers (Table 1) were removed in these trials as they could interfere with the swing motion. During data processing, the static trials were registered and the secondary marker locations were computed automatically based on the static trials and the primary markers, including the ball, clubface points, and joint centers.