An image processing apparatus includes a CPU and an imaging unit. In this image processing apparatus, an image captured by the imaging unit is displayed on an LCD together with an imaging index, indicative of a predetermined region of a face. In response to a shutter operation by an operator, a portion, matching with a position of the imaging index, is extracted from the image data as a region image. Furthermore, in response to a shutter operation, another image captured by the imaging unit is stored as a still image. Then the region image and the still image are composed with each other to display a composite image, obtained by pasting the image of the predetermined region of the face on to another image. In addition, with respect to the composite image, the region image is deformed according to a feature of the data from the imaging unit, an elapsed time without an operation input, and so on.