Research of the approaches to construction of systems of automatic reading of the symbolical information.
The given work is devoted to development of the prototype of algorithm and software for information systems of automation of document circulation. The basic principles of construction of automatic reading system are considered. The original algorithms of detection of textual lines, determination of symbol places and adaptive binarization of the images are offered. Method of zones and method of crossings are implemented for recognition of OCR-B-code symbols.