A Local Fingertips Movement and Fingertips Clustering Based Virtual Keyboard Adopting a Camera
Hui Ji, Jianxin Chen, Qingyu Lin, Ang Li · 2018
With the development of virtual reality technology, human-computer interaction (HCI) changed from the traditional keyboard, mouse into many virtual methods. Different from the traditional virtual keyboard system recording and processing the all fingertips, our system pays attention to the keys been stroked, which means we only process the fingertip coordinates located in the key region. In addition, we also propose a fingertip extraction algorithm based on contour feature restriction and coordinate clustering. Compared to other virtual keyboard system, our system is simple but still effective. The extensive experiments demonstrate that our virtual keyboard system is flexible (only requires the camera in front of and above the keyboard to make sure the whole panel is detected by the camera), efficient (15 FPS on average, robust to cluttered backgrounds and placement between the camera and the paper keyboard, accurate (a 95.20% mean accuracy on 59 common used keys).