期刊名称：International Journal of Hybrid Information Technology
印刷版ISSN：1738-9968
出版年度：2015
卷号：8
期号：10
页码：141-152
DOI：10.14257/ijhit.2015.8.10.13
出版社：SERSC
摘要：As mobiles become cheaper and gain popularity, mobile applications are developing quickly and trying to utilize more techniques to make these applications more efficient and usable. Augmented reality is a live direct or indirect view of a physical, real-world environment whose elements are augmented (or supplemented) by computer-generated sensory input such as sound, video, graphics or GPS data. This research develops an android mobile application that utilizes real-time image processing techniques to provide simulated functionality to a hand drawing of piano keys. A piano keyboard with fourteen white keys and ten black keys is drawn used a black thick pen on paper, and the mobile is set in an angel that gives the optimum view of the piano keyboard when the camera is turned on. When the application is started, a camera shot of the piano, without the presence of hands or fingers in the view, is taken. The user can then start playing on his drawn piano. The application detects keystrokes and decides which key is pressed; and then plays the corresponding tone. The application is developed using OpenCV library which focuses on real-time image processing and computer vision, since the target devices with android platform, a Java program is used to port the C++ code to android. The application is tested on multiple android-based devices with different specifications under good lighting conditions. The empirical results are impressive and comparable across different devices, despite all changes in lighting and background; all devices exhibited the same level of accuracy in detecting fingers and drawn pianos.
