An image processing apparatus includes: a luminance extraction means to extract luminance components of an input image; a contrast extraction means to extract contrast components of the input image based on the luminance components of the input image extracted by the luminance extraction means; a storage means to store a performance function indicating relation between the contrast components of the input image and depth amounts subjectively perceived, which is determined based at least in part on visual sense characteristics of human beings; and a contrast adjustment means to calculate present depth amounts of the input image from the contrast components of the input image extracted by the contrast extraction means based at least in part on the performance function with respect to region of the input image which are determined from depth information of the input image and adjusting contrast components of the input image.