The topic of this master’s thesis is learning of motion patterns, which is also known as motion learning or motor learning. The progress of such learning is based on feedback on the performance of a motion pattern, on the basis of which a student can improve his performance in the following repetitions. The feedback can be internal or external. While the first originates from human’s internal perception processes (so-called proprioception), the external feedback (EFB) originates from human’s external environment and is added to the internal feedback with the purpose of making the motion learning process even more effective. The EFB can be transmitted through sound, visual, or haptic perceptual channel, or in a combination of them. In the thesis, we deal with visual EFB, implemented in the form of web application. Web applications are implemented in browsers. Due to their prevalence web applications are generally useable on various devices.
Motion patterns are classified with regards to their complexity, cyclicality, and the speed of their performance. All these characteristics influence the strategy of providing the EFB. The latter includes the decisions on numerous characteristics of EFB, such as the used perceptual channel, the source and the moment of providing EFB, the freedom of an individual’s decision on EFB, polarity, and the difficulty of processing. Each of the mentioned characteristics is processed in the corresponding subchapter in more detail.
In continuation, we focus on visual feedback, which can be presented in natural or in abstract forms by means of various strategies. In the case of visual EFB, the choice of a form and strategy depends mostly on the complexity of learned motion patterns. The record of such pattern consists of accurately listed data of value of so-called motion variables, which describe the motion. They are the result of the so-called motion capture, which in the master’s thesis is performed by means of the system for optical motion capture. Motion variables have to be recorded in the appropriate data structure, suitable for further analysis of motion patterns and preparation of the EFB. The data structure, defined within the framework of the master’s thesis, enables recording the values of any motion variables in space and time.
In the practical part of the master’s thesis, we focused on depicting the EFB by means of web technologies. Since direct access to graphic hardware is not possible nowadays, nor it makes sense, the so-called Application Programming Interfaces (APIs) for graphics display are used for rendering images. We use WebGL (Web Graphics Library) API for rendering in the web application. WebGL is designed especially to be used on the web. Rendering is performed in the client part of the web application. The client is provided data on motion by the server part relying on communication via WebSocket protocol. With the help of the latter, the data on motion is transferred to the web application in real-time, which means that a student can receive EFB at the moment when he or she completes performing the motion pattern.
Due to the predefined data structure which serves as an access point to the service of the developed application, the application is transferable and does not depend on an actual motion capture system but is applicable more widely. The feedback provided is intended for general motion (is not limited only to one type of motion) and is displayed either abstractly or in virtual environment (the so-called natural display), or by a combination of both. The use of the developed web application and more detailed description of the supported visualizations are presented in the example of a golf swing.
|
