Simplified Initial Setting of a Nurse Call System Driven by Eye Movement Using Image Processing

抄録

In this research, a nurse call system with simplified initial setting is developed. The system is constructed by improving an existing nurse call system driven by eye movement, where the movement is detected by image processing. In the initial setting of previous research, template images for the detection of face and eye were created manually. However, it was not easy for system operator to generate suitable template images. To solve this problem, in the research, automatic generation of template images is attempted to simplify the initial setting. The system with simplified initial setting is realized by introducing gamma correction, cascade classifier and false positive rejection. The usefulness of the proposed system was verified through operation of not only initial setting but also nurse call.