Methods and apparatus for automatic modification of semantically significant portions of a document without document image decoding

Abstract From the Canadian patent

Methods and apparatus of processing an undecoded document image in a digital computer to modify the document image so as to emphasize semantically significant portions without first converting the document image to character codes. The document image is segmented into image units, and morphological image characteristics of the image units are evaluated to identify significant image units for emphasis. In one embodiment the , significant image units are emphasized by modifying at least one shape characteristic of the significant image units using at least one uniform morphological bitmap operation applied to the entire image unit bitmaps corresponding to the significant image units.