Document retrieval method and apparatus using image contents

Abstract

The invention provides a document retrieval method and device based on content images. Conventional document retrieval methods need to read the document at a time in order to consider the content, so far it has not been possible to realize the at-a-glance content. Therefore, although there is a method of generating a summary image in advance, it takes a lot of time and cost to generate it. In addition, because there are Because it is difficult to display a summary image of all the documents retrieved on one screen of the monitor, the problem cannot be completely solved. The present invention replaces the document with an image with high visibility, that is, the photographs, diagrams, tables, etc. in the document The image is used as the key image of its document, and one or more retrieval logic formulas using its key image are input, and all images in the document containing similar images satisfying the retrieval logic formula are three-dimensionally displayed on the display screen, and if the searcher chooses If the key image for is displayed, the document content including its key image is displayed.

Description

technical field [0001] The present invention relates to a method and device for retrieving documents using word-processed images. In particular, it relates to an effective retrieval method and device for documents, and the documents mentioned here also include Web documents constructed on the WWW. In addition, the so-called images include photographs, screens, diagrams, tables, graphs, symbols, and other various contents appearing in documents. Background technique [0002] Conventionally, in specific fields such as patents and medical fields, it has been necessary to search past document examples from the viewpoint of novelty examination of patents and examination of similar cases, and is currently being actively researched and developed. On the other hand, recently, a search technology or software for obtaining information needed by individuals from the Internet or intranet with a well-established infrastructure such as a network is being developed. The mainstream of ret...

AI Technical Summary

Problems solved by technology

That is, in document retrieval, it is necessary to read documents at a time in order to consider the content, and it has not been possible to realize the content at a glance

For this reason, although there is also a method of generating a summary image in advance as in JP-A-05-216936, it takes a lot of time or cost to generate it, or it is difficult to display the summary images of all documents retrieved on one screen of a monitor, etc. Due to reasons, the problem cannot be completely solved

Furthermore, in general keywords, it is often not possible to effectively retrieve the required documents

This is because generally only according to the frequency of keywords in the document, it is often difficult to approximate its content correctly, so it is not possible to only detect articles with similar content

Images