Abstract
Signs are ubiquitous indoors and outdoors, which are used for way finding, finding shops and businesses, accessing variety of services. But the information of signs is inaccessible to many visually impaired people unless they are represented in a non-visual form such as Braille, tactile graphic, and speech. In this paper, we propose an approach to extract character string from scene images. In our method, we first segment the scene image into homogeneous regions by applying toggle mapping. Second, we detect the character edges from homogeneous region, after that, single characters are detected. A fuzzy inference system is applied to extract character strings. Finally, we conduct computer experiments to show the effectiveness for the proposed method.
