The aim of this bachelor thesis is to develop software for evaluating the performance of skateboarding tricks. Specifically, the goal is to design a data-processing algorithm and develop desktop application that can record and compare two performances of the ollie maneuver. Additionally, efforts were made to create a convenient and modern environment for creating applications using the Microsoft Kinect device.

The comparison algorithm was created using Python, and it involves two types of indicators that can be used to distinguish between two trick performances. The first indicator is based on the variations in timing between detected events occurring within the duration of the ollie, while the second indicator involves comparing time series representative of the angle between the skater’s hip bones (crotch angle). The comparison process begins with analyzing the recorded data in order to identify consecutive stages and events characteristic of the maneuver. The computed values of the indicators can then be used to post-process and present the desired results. Importantly, the proposed code solution is adaptable, and the indicators can be used to extract results from arbitrary stage.

The desktop application was developed using TypeScript programming language with Electron and React frameworks. It utilizes Microsoft Kinect v2 drivers to collect data from the depth sensor, and can display recording previews and comparison results thanks to popular web development technology stack including HTML, CSS, and React. This application can serve as a learning tool for skateboarders.