Browse Prior Art Database

A system and method to read meaningful table content in software with image analysis Disclosure Number: IPCOM000241633D
Publication Date: 2015-May-19
Document File: 5 page(s) / 131K

Publishing Venue

The Prior Art Database


In software, accessibility is very importance for user, especially when user is a low vision people. There is an item in accessibility standard to assist low vision people to learn the content of table component in software correctly. To implement it, it normally requires software developer to apply a couple of specially technique to table component. It will cost a lot of effort. This invention devises a method to learn table content using image analysis. The method can get the table content correctly without the restriction of current implementation – it can be widely used in web application and desktop application, it does not require software developer to do any addition effort to apply special technique, it can be enhanced to any existing screen reader software without confliction.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 44% of the total text.

Page 01 of 5

A system and method to read meaningful table content in software with image analysis

Accessibility is very important to all software including web application and desktop software. In Accessibility for 508, there is a check point that requires software to enables low vision user to learn the table content correctly with the assist of software like screen reader.

Currently, to make software to be compliance with accessibility requirement, it usually requires software developer to apply a couple of standard techniques to a table in software UI. However, it normally cost a lot of effort for a developer to do so. On the other hand, there are two cases in which it is hard for a developer to apply the standard technique. Case a, the table is from a third party software which does not follow up accessibility requirement itself, it will hard for a developer to enhance it. Case b, the table is constructed by a non-standard element. Take web for example, one table is constructed by a couple of DIVs, which is hard to apply the standard technique to it.

This invention proposes an easy way to recognize table content in software and read it for low vision person. The method introduces image analysis mechanism to recognize table header cell and content cell, collects each cell text and re-organizes the texts together to make them meaningful to low vision people.

As far as we could see, the advantage of the invention is obvious.

1. It can reduce the effort for a software developer of applying standard techniques to software to meet accessibility standard .

2. It can solve the problem when a using third party software which is not accessible itself.

3. It can make a non-standard table to be accessible.

4. The method has no conflict with existing screen reader's mechanism but can be enhanced into it as an alternative function - e .g.

   a screen reader will use the proposed function while it cannot read a table content correctly.
5. The invention can be used in both web-based software and desktop software.

This invention is used to recognize table content in UI page and reconstruct recognized table content into meaningful text which could be understood by low vision person. The Figure 2 is the whole processing flow to extract meaningful text from any given UI page which contains table. It contains four major steps.


Page 02 of 5



Figure 2 - the processing flow of generating meaningful text from a table

111::: Snapshot an UI page into picture

Snapshot an UI page into picture

This step of process just needs to use any existing technology to snapshot a UI page into a picture. It is a simple step, but it is necessary for Table Recognition by Image Analysis.



222::: Table cell recognition by Image Analysis

Table cell recognition by Image Analysis

In this step, UI page has been captured by snapshot. This step is the core claim point of the invention. Imagine the picture contains a table like the following Figure 3 shows.


Page 03 of 5