An image processing apparatus specifies, based on a reference image out of a plurality of images and a plurality of comparative images out of the plurality of images, a difference region, in each of the plurality of comparative images, including an object subjected to combination corresponding to a difference from a reference image, determines, based on a plurality of difference regions specified in the plurality of comparative images, an object region corresponding to an object included in the reference image, and combines, based on the determined object region in the reference image and the plurality of difference regions in the plurality of comparative images, with the reference image, the objects subjected to combination included in the plurality of difference regions so that an object corresponding to the object region is included in the reference image with which the plurality of difference regions are combined.