Posted: Fri 9 October 2009

Document OCR with Google Docs

Google are intergrating OCR software into Google Docs which means you can upload a scanned image and Google will pick out the text and turn it into an editable format within a document. Why is this interesting? Well it is interested because of how advanced the software is – the software has a very good understanding of what is text within an image and a very good ability to understand the specific words. This blog shows an example of a scanned image and the editable text that is generated from it.

This is an important advancement because if Google has software that allows it to understand and contextualise images then a big flaw in the search engine intelligence is overcome. Currently Google can’t read images, if you have images on a page then Google has to use some code in order to figure out what the image is about – this code is called “alt text”. The fact that the search engines have to use representation code that the user doesn’t see instead of analysing the image that the user does see has long been an issue for search engines. Now Google has software to analyse and understand an image so it no longer needs alt text and when this software gets rolled out to their search engine you will see more weight being given to images within website and also you are likely to see a mixing up of the search engine results.

There is no guarantee when this will get included in the natural rankings but you can be sure that if Google are testing new functionality that can have a benefit to the way they rank websites then they will use it.

By: Andrew Gaukrodger
http://www.eazytiger.net
Web design and ecommerce Leicester

||||

• « back to articles