Sonification of images for the visually impaired using a multi-level approach

This paper presents a system that strives to give visually impaired persons direct perceptual access to images via an acoustic signal. The user explores the image actively on a touch screen and receives auditory feedback about the image content at the current position. The design of such a system involves two major challenges: what is the most useful and relevant image information, and how can as much information as possible be captured in an audio signal. We address both problems, and propose a general approach that combines low-level information, such as color, edges, and roughness, with mid- and high-level information obtained from Machine Learning algorithms. This includes object recognition and the classification of regions into the categories "man made" versus "natural". We argue that this multi-level approach gives users direct access to what is where in the image, yet it still exploits the potential of recent developments in Computer Vision and Machine Learning.