An image processing apparatus is provided that reduces a data size of a composite file without affecting output when generating a composite file by merging multiple files containing objects. To accomplish this, in merging multiple files, the image processing apparatus determines whether or not objects (images or the like) contained in the multiple files are hidden behind another object. Further still, the image processing apparatus merges objects that are hidden behind another object as link information in the composite file, and image data corresponding to the objects is stored in a separate file.