A Quantitative Model To Evaluate Wrist-Rotation In Golf (P7)
Use of inertial sensors in BSNs is motivated by biomedical applications such as fall detection, gait analysis, sport medicine and balance assessment, and has received much attention during recent years. Authors and citeghasemzadeh2009eei introduce a framework for human action recognition using motion sensors. They integrate on-body sensors including accelerometers and gyroscopes in a wireless sensor network to classify physical movements. Barnes et al. perform a preliminary study on the effectiveness of body sensor networks for locomotion monitoring. Logan et al. report the results of a study on activity recognition using differentty pesofsensory devices, including built-in wired sensors, RFID tags, and wireless motion sensors. Maithe et al. use a tri-axial accelerometer mounted on the waist to recognize basic daily movements using a hierarchical classiﬁcation scheme. A pattern recognition technique for evaluating the performance of the human postural control system using inertial and EMG sensors is presented.
Advances in technology has enabled design of sports feedback systems which accelerate training by providing students with information regarding mistakes made during practice. Spelmezan et al. present an on-body wireless sensor platform for realtime snowboard training. They deploy inertial sensor, bend sensors and force-sensitive resistors along with communication facilities in a wireless network to capture and analyze rider’s motion and posture on the snowboard.K wonetal. develop amotion training system by integrating on-body accelerometer and motion capture data to detect human action and provide visual feedback in real-time. A motion capture system is used to design a virtual baseball batting training system where batter swings the bat toward a virtual ball rendered over a screen. The trajectory of the swing is then used to provide qualitative results. Golnalez and Alvarez study the problem of step estimation which is an important issue in designing coaching systems. Despite their successful develop ment of sports training systems, unique complexities in the golf movements make the aforementioned techniques inappropriate for a golf swing trainer.
A variety of golf swing analysis aids have become popular recently, making use of technologies such as high-speed photography, inertial sensors, and motion capture using magnetic, radio frequency, or ultrasonic markers. These systems can incorporate either devices placed in the sports environment or sensor sembedded with in the sport sequipment and human body. Urtasun et al deﬁne a temporal motion model that allows them to accurately extract 3D golf swing motion from a single camera while no markers are required to be placed on the subject.
The model allows them to overcome the obstacles of subject self-occlusion and movements which are perpendicular to the camera plane. This approach simpliﬁes the computational complexity typical in motion capture by using only one camera instead of many. Betzler et al describe the application and limitations of 3D motion analysis in measuring golf swing motion. Golf-speciﬁc limitations of 3D motion analysis include the high velocity of the hands, club and ball; inaccuracies in determination of body segment rotations; vibration of markers at impact; and marker placement and occlusion. The authors designed a test setup to minimize the effect of these errors by using cameras and careful placement of several markers on the golfer and golf club. Kiat present a different approach to measure golf-swing motion. They place electrogoniometers along the left arm and utilize dual Euler angles and dual Euler velocity analysis to estimate the location and velocity of the club head.