A personal computer issues an instruction to estimate a subject included in an image represented by image data stored in an image server. A database in which a place of photography, time periods, and a photographing direction correspond to structure image data representing a structure is stored in a storage unit of a subject estimating device in the image server. Related structure image data related to the place of photography and the photographing direction and a date/time of photography, are extracted from the database by referring to the position data and the date/time data attached to the image data. Then, a correlation value between the extracted related structure image data and the image data is obtained, and a structure included in an image represented by a related structure image data with the highest correlation value is estimated to be a subject included in the image represented by the image data.