Character reading system
First Claim
1. A method of reading characters of the type including the steps of scanning a character with a sensor having an array of photosensitive elements so that the illuminated or nonilluminated elements in said array correspond to segments of said character which collectively form an image of said character and processing the output signals from said elements to thereby identify the scanned character, wherein the improvement is characterized by said processing step comprising:
- determining the character area of said array in which said image is formed;
analyzing the output signals from each row or column of the elements within said character area and assigning a true classification value to each row or column representative of the pattern of illuminated and non-illuminated elements therein;
analyzing the sequence of at least some of said true classification values in order to identify said character; and
detecting the presence or absence of a character segment in a column of elements near one side of said character area to thereby determine the direction of movement of said scanner relative to said character.
0 Assignments
0 Petitions
Accused Products
Abstract
An optical character reading system first determines the height and width of the character to determine its authenticity and then examines the features of each line and/or column and classifies each line or column according to its pattern. The sequence in which the classification numbers occurs is indicative of the character detected. Correction of oblique scanning and direction determining features are also included.
24 Citations
12 Claims
-
1. A method of reading characters of the type including the steps of scanning a character with a sensor having an array of photosensitive elements so that the illuminated or nonilluminated elements in said array correspond to segments of said character which collectively form an image of said character and processing the output signals from said elements to thereby identify the scanned character, wherein the improvement is characterized by said processing step comprising:
-
determining the character area of said array in which said image is formed; analyzing the output signals from each row or column of the elements within said character area and assigning a true classification value to each row or column representative of the pattern of illuminated and non-illuminated elements therein; analyzing the sequence of at least some of said true classification values in order to identify said character; and detecting the presence or absence of a character segment in a column of elements near one side of said character area to thereby determine the direction of movement of said scanner relative to said character. - View Dependent Claims (3, 4, 5, 7)
-
-
2. A method of reading characters of the type including the steps of scanning a character with a sensor having an array of photosensitive elements so that the illuminated or non-illuminated elements in said array correspond to segments of said character which collectively form an image of said character and processing the output signals from said elements to thereby identify the scanned character, wherein the improvement is characterized by said processing step comprising:
-
determining the character area of said array in which said image is formed; assigning logic values to the illuminated and non-illuminated elements in each row or column of the elements within said character area so that each row or column is represented by a first binary sequence in which a "1" corresponds to a character segment; computing for each said row or column a second binary sequence comprising the logical sum of the first binary sequences of said row or column and the previous adjacent row or column; assigning first and second classification values to said first and second binary sequences; selecting one of said first and second classification values as said true classification value so that character identification errors caused by oblique scanning of said characters may be corrected; and analyzing the sequence of at least some of said true classification values in order to identify said character. - View Dependent Claims (6)
-
-
8. A character reading system of the type in which characters are scanned by a sensor having an array of photosensitive elements so that the illuminated or non-illuminated elements in said array correspond to segments of said character which collectively form an image of said character and the output signals from the elements are processed to thereby identify the scanned characters, wherein the improvement comprises:
-
means for determining the character area of said array in which said image occurs; means for analyzing the output signals from each row or column of the elements within said character area and assigning a true classification value to each said row or column representative of the pattern of segments therein; means for analyzing the sequence of at least some of said true classification values in order to identify said character; and means for detecting the presence or absence of a segment in a column of elements near one side of said character area to thereby determine the direction of movement of said scanner relative to said character. - View Dependent Claims (10, 12)
-
-
9. A character reading system of the type in which characters are scanned by a sensor having an array of photosensitive elements so that the illuminated or non-illuminated elements in said array correspond to segments of said character which collectively form an image of said character and the output signals from the elements are processed to thereby identify the scanned characters, wherein the improvement comprises:
-
means for determining the character area of said array in which said image occurs; means for assigning a first classification value to each row or column of elements within said character area based on the pattern of segments occurring therein; means for logically summing the pattern from each row or column with the pattern from the previous row or column; means for assigning a second classification value to said row or column based on the pattern of segments in said combined pattern; means for selecting one of said first and second classification values as said true classification value; and means for analyzing the sequence of at least some of said true classification values in order to identify said character. - View Dependent Claims (11)
-
Specification