Method for improving the accuracy of character recognition processes
The present invention embodies a character recognition method for constructing a result string from a plurality of result sets. Each result set comprises at least one candidate character, and each candidate character has an associated confidence indication. The method can begin by selecting a plurality of character types. For each selected character type, a candidate string can be created by concatenating a candidate character of the selected character type from each result set. The associated confidence indication for each concatenated candidate character can be combined to form a corresponding combined confidence indication for each created candidate string. The created candidate string with the most favorable corresponding combined confidence indication can be selected as the result string.
1. A character recognition method for accurately constructing a result string, the method comprising the steps of:
generating a digital representation of an input string including characters;
generating a result set for each character in the input string, each result set including a plurality of candidate characters and a plurality of associated confidence indications;
selecting a plurality of character types;
for a first selected character type, creating a first candidate string by concatenating candidate characters with a most favorable corresponding confidence indication of the first selected character type from each result set;
for a second selected character type, creating a second candidate string by concatenating candidate characters with a most favorable corresponding confidence indication of the second selected character type from each result set;
if there is no candidate character for the first candidate string in a result set for the first selected character type, extracting a substitute candidate character of the second selected character type;
concatenating the substitute candidate character with candidate characters with a most favorable corresponding confidence indication of the first selected character type from each result set to thereby form the first candidate string;
ascribing a substitute associated confidence indication for the substitute candidate character;
for the first and second candidate strings, combining the associated confidence indication for each concatenated candidate character to form a corresponding combined confidence indication; and
selecting as the result string the created candidate string with a most favorable corresponding combined confidence indication.
2. The method of claim 1 , wherein the substitute candidate character corresponds to a most favorable corresponding confidence indication of the second selected character type for the result set.
3. The method of claim 1 , wherein:
each confidence indication is numeric; and
the most favorable corresponding combined confidence indication has a greatest combined value.
4. The method of claim 1 , wherein the first selected character type includes a plurality of subset character types.
5. The method of claim 4 , wherein the first selected character type further includes a pattern for positioning the candidate characters of the plurality of subset character types.
6. A computer readable medium having computer instruction code for performing a method of character recognition to accurately construct a result string, the method comprising:
generating a digital representation of an input string including characters;
generating a result set for each character in the input string, each result set including a plurality of candidate characters and a plurality of associated confidence indications;
selecting a plurality of character types;
for a first selected character type, creating a first candidate string by concatenating candidate characters with a most favorable corresponding confidence indication of the first selected character type from each result set;
for a second selected character type, creating a second candidate string by concatenating candidate characters with a most favorable corresponding confidence indication of the second selected character type from each result set;
if there is no candidate character for the first candidate string in a result set for the first selected character type, extracting a substitute candidate character of the second selected character type;
concatenating the substitute candidate character with candidate characters with a most favorable corresponding confidence indication of the first selected character type from each result set to thereby form the first candidate string;
ascribing a substitute associated confidence indication for the substitute candidate character;
for the first and second candidate strings, combining the associated confidence indication for each concatenated candidate character to form a corresponding combined confidence indication; and
selecting as the result string the created candidate string with a most favorable corresponding combined confidence indication.
7. The computer readable medium of claim 6 , wherein the substitute candidate character corresponds to a most favorable corresponding confidence indication of the second selected character type for the result set.
8. The computer readable medium of claim 6 , wherein:
each confidence indication is numeric; and
the most favorable corresponding combined confidence indication has a greatest combined value.
9. The computer readable medium of claim 6 , wherein the first selected character type includes a plurality of subset character types.
10. The computer readable medium of claim 9 , wherein the first selected character type further includes a pattern for positioning the candidate characters of the plurality of subset character types.