Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Zonal Tree Features for Character Recognition

IP.com Disclosure Number: IPCOM000042647D
Original Publication Date: 1984-Jun-01
Included in the Prior Art Database: 2005-Feb-04
Document File: 2 page(s) / 38K

Publishing Venue

IBM

Related People

Bednar, GM: AUTHOR [+3]

Abstract

Character image data is tested progressively by a series of binary tests forming a decision tree. The tests continue until a particular character of a defined set of characters has been unambiguously identified. Each of the tests can be made on individual image bits or on collective image data known as measurements or features. Features generally represent characteristic shapes associated with individual characters. Features can themselves be identified by decision trees that test for predetermined image data patterns occurring in particular regions or zones of the character image. Such a feature decision tree consists of a progressive series of binary tests made on data selected from a defined region of the character image.

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 99% of the total text.

Page 1 of 2

Zonal Tree Features for Character Recognition

Character image data is tested progressively by a series of binary tests forming a decision tree. The tests continue until a particular character of a defined set of characters has been unambiguously identified. Each of the tests can be made on individual image bits or on collective image data known as measurements or features. Features generally represent characteristic shapes associated with individual characters. Features can themselves be identified by decision trees that test for predetermined image data patterns occurring in particular regions or zones of the character image. Such a feature decision tree consists of a progressive series of binary tests made on data selected from a defined region of the character image. The output of the feature decision tree is a binary indication of whether or not the feature is present. A simplified example of a decision tree to determine the feature "Flat Top" in the upper portion of the character image is shown in the illustration. At node 10, picture element A,2 is tested to see if it is black (yes) or white (no). If yes, then picture element A,3 is tested at node 11, and so on. As a result of the decision thus made, a single storage element is turned "on" or "off" to indicate the presence or absence of the feature "Flat Top". This storage element can itself be tested as part of the overall recognition tree logic.

1

Page 2 of 2

2

[This page contains 3 pictures or other non-...