The objective of this research is to design and prototype a low-cost optical character recognizer (OCR) with voice output for aiding visually-impaired persons to read books, magazines, and other printed materials. A primary objective is to design an easy-to-use device for a $1,000-$2,000 retail price. Ways of reducing costs are design of a hand-operated machine, relaxing speed and accuracy requirements, and design with low-cost parts. The principal methodology used will be developing the OCR algorithm with an OCR development workstation. The workstation features rapid prototyping of segmenting and recognition strategies, and a graphical user interface. The workstation has some of the features of the Small Talk artificial intelligence development system, and it is expected to speed up development and testing of OCR algorithms by a factor of four. A potential commercial application is to allow the visually impaired to be more productive. Additional commercial applications are graphics scanning and OCR input. The wider commercial applications are expected to provide incentives to lower the price to visually-impaired individuals. To this end, the proposer has located sources of private funds.
