摘要 |
PURPOSE:To obtain a detection method achieving a human three-dimensional position detection by a one-eyed input system. CONSTITUTION:The image of a space where presence of a human being is predicted is picked up, a region of a brightness value within a specified range corresponding to a human hair portion is extracted from the image, and then the region is estimated to be the human hair portion when the upper edge of the extracted region is in the form of a bow. Assuming that the horizontal width of the region which is estimated to be the hair portion is constant, the distance from the horizontal width of the image to the human being is calculated. |