Eye position detecting apparatus

摘要

An image of a face including eyes is converted into a binary coded image. The binary coded image is processed to detect first and second rectangular regions including image portions representing the eyes, respectively. For this purpose, the binary coded image is scanned in a series of horizontal scan lines to detect the X and Y coordinates of initial and final pixels of a line segment representing the longest width of the face. The X coordinates of each of the first and second rectangular regions are calculated based upon the detected X coordinates of the initial and final pixels. An image portion above the line segment is scanned to detect the Y coordinate of a lowermost one of pixels forming black image portions. The Y coordinates of the first and second rectangular regions are calculated based upon the detected Y coordinate of the lowermost pixel.