<![CDATA[In the computer science both the operating system and the display during installation, there is a feature called a menu. This feature used to access functions on the running system. Has 2 basic display are GUI and CLI, most menus use human hand for navigation. The problem is that the compilation menu can't be accessed by persons with disabilities and person who get accident that cause a problem in their hand to move an object or hardware for navigation a menu. The solution offered uses detection of face edges using human head with raspberry pi single board. RGB convert to YCbCr and HSV color space, then forming a bounding box with parameters x, y, w, h and also quadrant area . By determining the value of h / w and centroid distance and the number of pixels in each quadrant as a determinant of movement classification. The results obtained based on lux 1500-1800 reached 96% and lux 1250-1400 reached 92% while for distances at 40,50,60 and 80cm had the highest accuracy at 60cm distance with 96% accuracy . With a good result in visual display and oval shape is also good in accordance with the expectations of researchers.]]>
Copyrights © 2019
