Abstract: A television doorphone apparatus in which an image of a quest is picked-up by a television camera provided on an outer wall of an entrance of a house, an image signal generated by the television camera and representing a whole image is processed to generate an enlarged image including the head of the guest, and the enlarged image is displayed on a monitor screen. In order to produce the enlarged image, at first a top center of the head of the guest is detected by comparing a relevant detection image with a previous detection image, the detecting image being constructed by smaller number of pixels than the image signal. Then, a start address of the enlarged image having a smaller size than the whole image is derived from the detected top center position of the head of the guest. Next, the pixel data of the enlarged image is extracted from the image signal, and finally the extracted enlarged pixel data is interpolated.