In order to easily and rapidly determine a subject region in a captured image using a single image, a synthesized image of an extraction image including a subject and an instruction image representing a first acquisition region from which color information on the subject is acquired is used. Color information on the subject is acquired from the first acquisition region represented by the instruction image in the synthesized image. Color information on a background of the subject is acquired from a region not including the first acquisition region and a color information non-acquisition region that is adjacent to the first acquisition region and is set in advance in the synthesized image. On the basis of the color information on the subject and the color information on the background, extraction information of the subject is determined and output.