An image processing apparatus includes: an acquisition unit configured to acquire band images of an object, using filters of which transmission center wavelengths are different from each another; a detection unit configured to select a first band image from the band images, and detect a target region that is to be improved with respect to visibility; a selection unit configured to select a second band image, which includes information, originated from a structure of the object, within the target region, from among the band images other than the first band image; and a generation unit configured to generate a spectral image from a band image. The generation unit uses at least a band image that has been captured using a same filter as the second band image, upon generating a spectral image of the object.