Fingertip detection for virtual keyboard based on camera

Tablet producers offer many utilities for their tablets to attract the users. The virtual keyboard based on camera can be such a utility in the near future. We have designed the strategy for an estimation of the finger positions in a virtual keyboard space and detection of clicking events. This strategy includes the processes: picture capturing, edge detection (finger contour detection), fingertip localization, fingertip tracing and click detection. A key step in the whole strategy is the exact localization of fingertip. In this article we describe in detail three proposed solutions to this task. The first algorithm is based on finding local maxima of the finger contours, and this maximum has to meet certain criteria. The second proposed method uses a curvature function calculated from a traced hand contour - a curvature function has the local maxima at the fingertips. The third method is based on angle between two vectors created from a traced hand contour - angle has a minimum at the point of fingertip. In our article we describe the principles of the three proposed methods as well as the practical experiences that we gained in real conditions.