Systems and methods for image deduplication in a file are provided. A method includes identifying a plurality of similar images (e.g., similar intent images) in an electronic document. The method also includes designating one of the plurality of similar images as a master copy. The method also includes replacing each remaining one of the plurality of similar images, other than the master copy, with one of (i) a duplicate copy of the master copy and (ii) a soft link pointing to the master copy.