An image processing apparatus includes a reading unit and a correction information output unit. The reading unit acquires a saddle-stitched surface picture by reading a saddle-stitched surface of a saddle-stitched booklet formed through folding and stapling processes performed by a saddle-stitching device, and acquires a cover surface picture by reading a cover surface of the saddle-stitched booklet. The correction information output unit outputs correction information about a stapling position to the saddle-stitching device. The correction information is calculated on the basis of a difference between a cover-side fold line position on the cover surface picture, which is specified on the basis of a saddle stitch-side fold line position detected from the saddle-stitched surface picture, and a staple position detected from the cover surface picture.