An image processing device 1X includes a three-dimensional reconstruction means 31X, a matching means 32X, and an output control means 33X. The three-dimensional reconstruction means 31X generates reconstruction data Mr acquired by three-dimensionally reconstructing an inspection target on a basis of captured images Ic which a photographing unit provided in an endoscope generates by photographing the inspection target. The matching means 32X performs matching between a standard model “Ms”, which is a standard three-dimensional model of the inspection target, and the reconstruction data Mr. The output control means 33X displays on a display device 2X a photographed area indication image indicating a photographed area of the inspection target on the standard model based on a result of the matching.