Retrieval of relevant parts of document images based on density distributions of characters

摘要

This report presents a new method of document image retrieval that is capable of spotting parts of page images relevant to a user's Query. This enables us to improve the effectiveness and the usability of retrieval since the method is capable of spotting only relevant parts and thus free from the influence by irrelevant parts. Th( proposed method is based on the assumption that parts of page images which densely contain characters in a query are relevant to it. The characteristics of the proposed method are as follows: (1) Two-dimensional density distributions of a Query are calculated for ranking parts of page images, (2) The method relies only on the distribution of characters in page images so as not to be severely affected by the errors of character recognition and layout analysis. Based on the experimental results of retrieving Japanese newspaper articles, it is shown that the proposed method is superior to a method without the function of dealing with parts.