Discussing artificial intelligence is in itself a rather difficult undertaking in consideration of the vastness of the fields this discipline is linked to. It is impossible to write exhaustively on the subject in such a short article. What we will try to do in this article is to provide some insights and practical situations in which this science has found excellent application and others of interest for experts. First of all, let us try to give a definition albeit simplistic, of what artificial intelligence is: artificial intelligence is the discipline that deals with studying and measuring the ability of a computer hardware and software system to respond, as much as possible, as a human brain would in certain situations; the ability to independently make even complex decisions by examining the information stored in the system itself. A good example that has perhaps constituted the first area in which models of artificial intelligence have been conceived and realized to give answers of high degree of complexity is that of games; and more precisely chess and related games. In fact, for several decades, famous chess players at the international level have ventured into challenging computers programmed for this purpose with mixed results.
With this premise, let us examine one of the areas in which artificial intelligence has proved particularly useful and fruitful in terms of results achieved to provide greater autonomy to blind and visually impaired people in many aspects of their lives: from studies to the exercise of one's own work activity, from cultural education to the simplest and most immediate everyday needs; we are referring to the "treatment and processing of images." This theme is understood in the broadest possible sense. The problem and challenge are undertaken not only to eliminate the obstacles faced by people with disabilities, but to first and foremost respond to the need to digitize the printed materials that will inevitably disappear over time. The first aspect in the treatment and digitization of images was that of being able to extrapolate the written text by photographing paper supports. Brilliant and reliable solutions have been found. With the use of hardware devices for image acquisition and special OCR software (Optical Character Recognition), anyone can digitize printed books, magazines and newspapers in order to have the printed texts available for future use on an electronic device. Naturally, blind and visually impaired people have also benefited from the adoption of these technological solutions, since using assistive technologies to interact with electronic computer devices it is possible to use these texts extracted from printed materials with a margin of error that is almost non-existent.
One can only imagine the decisive impact that such a technological revolution has meant for people living with vision loss: almost thirty years ago, it took several months before a book was recorded or printed in Braille, so that blind and visually impaired people could have access to it. With the entrance on the market of the first scanners and OCR programs, the time before using a printed text has been drastically reduced from several months to a few hours! The Abbyy FineReader program for Windows and Mac OS platforms and the KNFB Reader App developed for Apple and Android mobile devices are among the most effective OCR applications in terms of processing speed and results. The second aspect of digitization and image processing on which are working the industrial giants in the IT field relates to the recognition of the content found in images so as to obtain textual descriptions of the images themselves. In short, they are trying to provide a verbal description of static images for the recognition and interpretation of the reality around us. In this regard, software platforms already exist which with the use of extremely complex algorithms are able to produce satisfying results both for facial recognition and the interpretation of texts. Of a certain importance is the platform that the well-known social network Facebook adopted to provide textual descriptions of the photos for anyone using assistive technologies (screen readers). In fact, using a screen reader software for computers or mobile devices, you will find that when browsing the pages of Mark Zuckerberg's social network when there are pictures, the Braille display or the voice synthesis used by blind and visually impaired people will almost always read a brief description of the contents of the photo. Not only Facebook has risen to the challenge: giants like Google and Microsoft are making important efforts in this regard. For visually impaired users, it is worth pointing out some apps, especially for mobile devices, which provide text and face recognition either by fishing between photos stored in a device, or by taking photos in real time: worthy of note are apps such as TapTapSee or CamFind, developed for both iOS and for Android which, using the computing power of a computer, perform image processing remotely returning textual descriptions on the mobile device where the image is found. Also, the app Seeing AI developed by Microsoft is a simple and intuitive interface which brings together many functions, including the immediate reading of texts from paper support or from the screen of an electronic device, the reading of handwritten text, the text and face recognition on pictures stored in your device or acquired in real time, the recognition of bar codes that identify products and even the detection of the degree of light in an environment. This last app performs the treatment of the image taking advantage of the very high-processing power now found on mobile devices. Seeing AI is currently only released for the iOS platform, and therefore for mobile devices made by Apple. At present, however, the results achieved are far from being considered acceptable. But the efforts made by the teams of scientists and software developers leave room for cautious optimism unthinkable just four or five years ago.