This paper describes the automatic index extraction algorithm for document database. First, the layout structure of the input documents is obtained in the storing process. And then the index information is extracted by matching the input layout structure with the document format information. This algorithm consists of the following three steps : (i) To recognize the category of input documents, (ii) To distinguish the index regions, (iii) To store the index information. Finally, the computer simulation results for the technical papers are presented.