A human face is extracted easily from an image including the face. An image including a human face and a face template comprising data of outlines representing a face shape are displayed on a monitor. A user carries out position matching between the face template and a face area which is an image of the face to be extracted from the image, by moving, transforming and/or rotating the face template through an operation of input means, for example. Extraction means calculates a degree of matching between the face area and the template according to the position matching operation. When the degree of matching exceeds a predetermined threshold value, notice of extraction of the face area is displayed on the monitor. The extracted face area is subjected to processing for converting a color tone thereof.