An image processing device includes a detector, a recognition unit, an acquisition, unit, and an output unit. The detector detects an image formed on a sheet, and a mark applied to the sheet. The recognition unit recognizes a mark position, which is a position where the mark is formed on the image. The acquisition unit acquires an additional image to be added and formed at the mark position. The output unit forms an output image using the image detected by the detector and the additional image acquired by the acquisition unit.