An image forming device that transfers a plurality of toner images of different colors in layers, onto an intermediate transfer member by controlling a timing of forming each of the plurality of toner images, based on a reference position on a surface of the intermediate transfer member, the image forming device comprising: an uneven part detecting unit operable to detect at least one uneven part that is any of a depression and a projection, on the surface of the intermediate transfer member; and a reference uneven part setting unit operable to set a reference uneven part that defines the reference position, from among the at least one uneven part detected by the uneven part detecting unit.