Computer-based system and method for character recognition

• US 5,214,719 A
• Filed: 02/28/1989
• Issued: 05/25/1993
• Est. Priority Date: 02/28/1989
• Status: Expired due to Fees

• Current Assignee: Imaging Associates Incorporated, Phoenix Imaging
• Original Assignee: Phoenix Imaging
• Inventors: St. Onge, James W., Budd, Gerald W., Krawec, Michael P.
• Primary Examiner(s): Goldstein, Herbert
• Assistant Examiner(s): COUSO, YON JUNG
• Application Number:
  US07/317,174
• Time in Patent Office: 1,547 Days
• Field of Search: 382/21, 382/25, 382/23, 382/19, 382/24, 382/42, 382/30, 382/13, 382/55
• US Class Current: 382/201
• CPC Class Codes: G06F 18/40 Software arrangements specially adapted for pattern recognition, e.g. user interfaces or toolboxes therefor

Abstract

A character recognition system and method is disclosed which provides for the teaching and recognition of characters. The system includes an apparatus and method for obtaining an image which includes the character, and for sampling the character to create a plurality of sample points. A vector correlation is performed on these sample points which compares the stored points to those of known characters. The system also provides a precedence table for determining the correct order of recognition of any correlation of multiple characters.

Claims

1. A method for recognizing a graphical symbol such as a character comprising:

(a) obtaining a digitized image which includes as a portion thereof said graphical symbol which is a character of an arbitrary height from a predetermined set of characters comprising a predetermined font;

(b) sampling at least that portion of the digitized image which contains said graphical symbol thereby creating a plurality of sample points at locations within the digitized image representing said graphical symbol;

(c) displaying said plurality of sample points;

(d) storing said plurality of sample points and their said locations within said digitized image, whereby said stored points and said locations are used to recognize said graphical symbol; and

(e) performing a vector correlation on said stored points and locations which compares said points with a set of known points from a known graphical symbol comprising a known character from said font to determine if said graphical symbol matches said known character; and

(f) using said vector correlation to determine a single correlation point for each graphical symbol.

2. A method for recognizing a character comprising:

(a) obtaining a digitized image using a camera which includes as a portion thereof said character;

(b) sampling at least that portion of the digitized image which contains said character thereby creating a plurality of sample points at locations within the digitized image representing said character;

(c) displaying said plurality of sample points;

(d) storing said plurality of sample points and their said locations within said digitized image, whereby said stored points and said locations are used to recognize said character; and

(e) performing a vector correlation on said stored points and locations which compares said points with a set of known points from a known character to determine if said character matches said known character by use of only a specified percentage of a total number of said plurality of said stored points.

3. The method of claim 2 wherein computer means is provided for storing a plurality of passwords and whereom, said method further comprising:

providing computer means for storing said specified percentage.

4. A method for recognizing a graphical symbol such as a character containing a plurality of pixels which may be activated and wherein said character is of a width equal to a number of pixels, the method comprising:

(a) obtaining an image digitized into a bit plane representation which includes as a portion thereof said graphical symbol;

(b) sampling at least that portion of the digitized image which contains said character thereby creating a plurality of sample points at locations within the digitized image representing said graphical symbol;

(c) displaying said plurality of sample points;

(d) storing said plurality of sample points and their said locations within said digitized image, whereby said stored points and said locations are used to recognize said graphical symbol; and

(e) performing a vector correlation on said stored points and locations which compares said points with a set of known points from a known graphical symbol to determine if said graphical symbol matches said known graphical symbol, and

wherein said sampling step (b) includes;

(1) activating said pixels contained within said character which intersect with the grid;

(2) reducing said width of said character to approximately one pixel; and

(3) positioning said grid over said reduced character; and

(4) defining said sample points to be the intersection of said grid with said reduced character.

5. The method of claim 4 wherein said sample points and wherein said method further comprises:

reducing said sample points to a width of approximately one pixel.

6. A method for recognizing a character wherein said character contains a plurality of pixels which are activated and wherein said character is of a width equal to a number of pixels, said method comprising:

(a) obtaining a bit plane image wherein said image includes said character upon a background;

(b) creating a field defining an area of said image upon said image wherein said field overlays said character and at least a portion of said background;

(c) activating pixels contained within said character to be readily distinguishable from pixels representing said background;

(d) reducing said width of said character to approximately one pixel in both horizontal and vertical directions;

(e) creating a grid over said reduced character such that said grid intersects said character at a plurality of points wherein said points are at a plurality of locations upon said character;

(f) displaying said character and plurality of points;

(g) storing said points and their said locations; and

(h) recognizing said character by means of said stored points and locations using vector correlation where only a specified percentage of a total number of said stored points need be matched with stored points and locations representing said character in prestored form.

7. The method of claim 6 wherein said method further comprising:

providing computer means for storing said specified percentage.