The classification of skateboarding trick manoeuvres through the integration of IMU and machine learning

The evaluation of tricks executions in skateboarding is commonly carried out subjectively. The panels of judges rely on their prior experience in classifying the effectiveness of tricks performance during skateboarding competitions. This technique of classifying tricks often fell short in providing accurate evaluations during competition. Therefore, an objective and unbiased means of evaluating skateboarding tricks is non-trivial. This study aims at classifying flat ground tricks namely Ollie, Kickflip, Shove-it, Nollie and Frontside 180 through the use of inertial measurement unit (IMU) and machine learning models. An amateur skateboarder (23 years of age ± 5.0 years’ experience) executed five tricks for each type of trick repeatedly on a customized ORY skateboard (IMU sensor fused) on a cemented ground. From the IMU data, a number of features were extracted and engineered. On the pretext of classification models, Support vector machine (SVM), k-NN, artificial neural networks (ANN), logistic regression (LR), random forest (RF) and Naïve Bayes (NB) was employed to identify the type of tricks performed. The results suggest that LR and NB have the highest classification accuracy with 95.0% followed by ANN and SVM together caped at 90.0% and RF and k-NN with 85.0% and 75.0%, respectively. It could be concluded that the proposed method is able to classify the skateboard tricks well. This will assist the judges in providing more accurate evaluations of trick performance as opposed to the subjective and conventional techniques currently applied.