Browse Prior Art Database

A method and system for searching highly related text based on uploaded image features

IP.com Disclosure Number: IPCOM000246542D
Publication Date: 2016-Jun-16
Document File: 3 page(s) / 46K

Publishing Venue

The IP.com Prior Art Database

Abstract

This is one method and system for searching highly related text based on uploaded image features. It extracts features from the uploaded picture, then outputs a word preference to search the related information in the text search engine. By it, it avoids the drawback of matching similar pictures which may be not highly related and problem of mismatching text and image in an article. In effect, users will get the exact results based on his uploaded ones, also much more information that he requests.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 3

A method and system for searching highly related text based on uploaded image features

Everyone may run into the similar situation that you find something interesting, then take photos and upload them to the search engine to check what it is. Maybe this is an antique, a beautiful flower, or a strange fish, etc., however, the search engine library probably only presents the similar pictures and the words who contains the similar pictures to you. But what you really need is the information about the picture you capture, not the information in the article who contains similar pictures. And what's more, the words in the article do not have the high relationship with the image. So, our disclosure proposes a new method to grab content information directly from the text search engine by the extracted characters of the uploaded picture.

This disclosure introduces a new method for searching information about the image the user uploaded. It extracts complex features from the uploaded image and then using these features to generate a words preference by analysis of natural language. At last, it utilizes this combined words preference to directly search image in the text search engine.

First, in this disclosure, it overcomes the obstacle that when there is no high related pictures in the image library. It extracts features from the picture to search the related information in the text search engine and avoid the drawback of matching similar pictures which may be not highly related. And second, as this searching method doesn't based on similar images, so this method will not e...