Browse Prior Art Database

Character Recognition

IP.com Disclosure Number: IPCOM000095602D
Original Publication Date: 1964-Mar-01
Included in the Prior Art Database: 2005-Mar-07
Document File: 1 page(s) / 11K

Publishing Venue

IBM

Related People

Baskin, HB: AUTHOR

Abstract

Many recognition techniques employ a buffer storage for the data that is obtained by scanning the character. The stored data is then analyzed to determine the identity of the character. When the buffer storage is sufficiently large to handle any character size which may be encountered, such as the upper case M or W in the largest anticipated type font, part of the storage is rarely used and, when used, contains data that is often non-discriminatory.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 1 of 1

Character Recognition

Many recognition techniques employ a buffer storage for the data that is obtained by scanning the character. The stored data is then analyzed to determine the identity of the character. When the buffer storage is sufficiently large to handle any character size which may be encountered, such as the upper case M or W in the largest anticipated type font, part of the storage is rarely used and, when used, contains data that is often non-discriminatory.

By using a buffer storage which is sufficiently large to accommodate most characters and by scanning oversized characters in segments, a large equipment reduction is obtainable at the expense of a small reduction in speed of operation. In many cases, even oversized characters are identifiable with the reduced size storage as there are sufficient discriminating features in the stored segments for unambiguous recognition.

Only in the case where ambiguity is present must the remainder of the oversized characters be interrogated.

1