| 摘要 |
<p>PURPOSE:To minimize the metal and physical burden on an operator by extracting an image of the eyeball part of the face of the operator, specifying picked-up pixels corresponding to the ellipse having the smallest flatness rate between a pair of ellipses approximating the eyeball part, and detecting a position where the operator stops. CONSTITUTION:When the operator looks at a display image on a display unit 1 to input and gazes at, for example, one left upper point, an image of the face of the operator is passed through fibers of respective pixels for image input and photoelectrically converted by a photodetecting element 6, and an image is composed by a phase composing circuit 7 and then inputted to an image recognizing circuit 9 through a signal processing circuit 8. The image recognizing circuit 9 extracts the parts of the eyes of the respective face images and approximates the irises of the right and left eyes with the pair of ellipses. The products of flatness rates of the respective ellipses are compared to detect pixels of the left upper part having the minimum flatness rate. Further, the positions of the detected pixels are inputted and the coordinates of pixels that the operator gazes at are generated and inputted to a computer 12.</p> |
