An insight into how the use of Inertial Measurement Unit sensors can play a role in shot classification in cricket batting
My research work at Griffith university gave an insight into how inertial sensors can be used in the game of cricket. Trialling these sensors firstly with bowling but completing my final report on batting. As researchers we looked at placing an inertial sensor on the thoracic vertebra to see what data could be recorded. This was the most practical and viable place to put a sensor as it can be used in a real game scenario. After using two elite batsman in manipulated and real game trials we recorded data onto a bought software to interpret it (Matlab). Our findings showed that when batsman’s recorded data in manipulated trials patterns emerged which could be identified and quantified. A sensor seemed to be able to detect when a ball was hit and trends on graphs could indicate what type of shot was played (shot classification). However, data recorded in a real game scenario was more problematic. Due to the numerous degrees of freedom in cricket batting it was problematic to interpret output graphs, especially when compared to the initial trials. In the future we hope that the sensors improve in order to record accurate and higher ranged data.
Overall, a big step was taken in this specific field and the idea holds huge viability in the game of cricket. Hopefully the next step is for my written up paper to be published and cricket teams become aware of the potential these sensors have in the future to record quick information elite players need when batting. Work on the IMU sensors at Griffith is ever increasing and improving to produce a exceptional IMU sensor.