An attempt towards the development of universal OCR U_OCR for the recognition of multi_lingual documents