Hand Fatigue Research On Computer Vision Based Gesture Interaction

Nowadays big-screen smart terminals are widely used in common life as the result of the rapid development of human-computer interaction(HCI) techniques. Gesture interaction based on machine vision has become one of the most high-profile branches of HCI techniques because of advantages such as simulated natural behavior, supporting for remote operation, allowing users to get rid of the shackles of interactive devices and so on. In recent years, technology research on gesture recognition accuracy and speed has become more mature and notable IT corporations have launched related products and applications. However, a large number of product reviews and user feedback show that fatigue on arm during interaction process has severely impacted user experience of using gesture interaction. Nevertheless, fatigue research of gesture interaction based on machine vision is still relatively scarce. Existing researches mostly focus on muscle strain while using gesture interaction, which results in the objectives and methods are difficult to be applied to the comfortable and natural gesture interaction, which makes it difficult to provide guidance for improving user experience via interaction design.According to theory of exercise-induced muscle fatigue on modern medicine and HCI research, this paper starts with biology theoretical basis of muscle contraction and fatigue, and then determines the subjective and objective union research methods. Combining of user-centered and bottom-up idea, fatigue caused by hover and move regional is explored according to real interaction circumstances. Furthermore, a model used for comprehensive fatigue assessment is put forward to predict user fatigue of gesture interaction system of different tasks and interaction circumstances. Real interaction circumstances have been built to prove its effectiveness. Chief and innovative work of this paper includes the following two parts:Firstly, this paper decomposes selection task of gesture interaction into dynamic process and static process because of the difference of muscle contraction between the two processes. Other than the existing research focusing on muscle strain, in this research, according to real interaction circumstances, muscle fatigue of the two processes is tested respectively by method of the control variables, which leads to the openness and universality of appliance.Secondly, this paper puts forward a model used for comprehensive fatigue assessment, which combines with changing trend of dynamic and static fatigue obtained in experiments above, system parameters of different task and empiric value such as moving time and human reaction time. Other than operating ISO-9241-9 experiments in traditional HCI research, this model is not used to evaluate interaction system by testing operation time and efficiency, but to put forward an innovative method to evaluate interaction by fatigue it caused, which could provide advice to interaction design.