Method and apparatus for summarizing a document without document image decoding

A method of excerpting significant information from an undecoded document image without decoding the document image, comprising:

• segmenting the document image into word image units without decoding the document image;
• deriving a word shape representation for each of a plurality of said word image units without decoding any characters making up said plurality of word image units, thereby deriving a plurality of said word shape representations;
• comparing said word shape representations to at least one other word shape representation to identify significant word image units from amongst said word image units;
• and
• outputting a plurality of said identified significant word image units for further processing.

A method for electronically processing an undecoded document image containing word text, comprising:

• segmenting the document image into word image units without decoding the document image;
• deriving a word shape representation for each of a plurality of said word image units without decoding any characters making up said plurality of word image units, thereby deriving a plurality of said word shape representations;
• comparing said word shape representations to at least one other word shape representation to identify significant word image units from amongst said plurality of word image units;
• forming phrase image units based on a plurality of said identified significant word image units, said phrase image units each incorporating one of said identified significant word image units and adjacent word image units linked in reading order sequence;
• and
• outputting said phrase image units.

An apparatus for automatically summarizing the information content of an undecoded document image without decoding the document image, comprising:

• means for segmenting the document image into word image units without decoding the document image;
• means for deriving a word shape representation for each of a plurality of said word image units without decoding any characters making up said plurality of word image units, thereby deriving a plurality of said word shape representations;
• means for comparing said word shape representations to at least one other word shade representation to identify significant word image units from amongst said plurality of word image units;
• and
• means for creating a supplemental document image based on said identified significant word image units.