A character region detecting unit detects respective character regions in a plurality of document images which are partitioned and read. A character recognizing unit recognizes the characters within the detected character regions, and converts them into character codes. An overlapping position detecting unit detects the positions of line images whose matching degrees are high as an overlapping position by making a comparison between the positions and the sizes of the character regions in the two document images. An image merging unit merges the two document images at the detected overlapping position.