Delivering graphical academic material such as diagrams, charts and graphs provides significant challenges for service providers. The Tactile Graphics Project ( TGP ) at the University of Washington firstly analyzed the work practices of tactile graphics specialists who create tactile forms of graphical images and proposed automated means by which to improve the efficiency of current work practices; and secondly, has described the implementation of various components of this new automated process.
We have developed a prototype Tactile Graphics Assistant (TGA) to enable specialists to more efficiently create tactile graphics one at a time and in batches. The TGA, using machine learning techniques, has the ability to identify the text in all the images in a given text book. The text can then be removed and converted to Braille. Using standard image processing tools the images, with the text removed, can be made more amenable to embossing by appropriate scaling and filtering. The TGA can then replace the original text with Braille in the embossible image. Our unique team of practitioners, researchers and consultants has contributed greatly toward achieving this goal. This presentation will describe the workflow involving the array of tools used to automatically produce graphics from a textbook in a batch process.
Prior knowledge of topic assumed: High
Contains advanced technical content: No
Back to the programme list
Back to the 2005 conference home page