期刊名称：International Journal of Advances in Soft Computing and Its Applications
印刷版ISSN：2074-8523
出版年度：2017
卷号：9
期号：1
页码：70
出版社：International Center for Scientific Research and Studies
摘要：This research proposes a head movement recognition method based on Android internal Gyroscope sensor. The contribution of this research is to proposes head movement as a new way of Human-Computer Interaction approach on Android smartphone. The head movement which has been used was an advanced movement from the single movement to the combination movement. Another challenge was head identification process based on built-in Android gyroscope sensor, which the sensor in the Android produced data noisier than the sensor in the iOS. Based on that, a preprocessing method has been proposed to enhance the head movement recognition. After the preprocessing phase, the head movement recognition phase conducted by the Nearest-Neighbour Classification Algorithm. Based on the proposed method, a controller application prototype has been made and deployed into the smartphone which combined with the dummy HMD to be used on the head. The experiment has been conducted on different hardware environments to understand the robustness of the proposed method. Based on the accuracy experiment, the success rate of our proposed method was about 95%. Based on the performance experiment, the average of the time amounts to execute the identification process was about 2.5 ms, which less than 100ms. Based on the experiment result, our proposed method is appropriate for a new experience of real-time human-computer interaction with head movement control only, although in a different hardware environment.
