In a method for forming an image of a subject on a solid-state imaging device, a first time period for splitting a light beam from a subject into a plurality of light beams that have different polarization directions and then combining the plurality of light beams to form a single subject image on the solid-state imaging device and a second time period for splitting the light beam from the subject into the plurality of light beams that have different polarization directions and forming a plurality of subject images that overlap each other partially on the solid-state imaging device are switched time-wise. A first image information on the single subject image is obtained based on pieces of signal information in the first time period, and a second image information on one of the plurality of subject images is calculated by using and computing pieces of signal information in the second time period. Then, the high-resolution image of the subject is achieved by using the first image information and the second image information. In this way, it is possible to obtain an image with a high resolution and a reduced noise with substantially no loss of the light beam from the subject.