System and method to recognise characters from an image
View Patent ↗System and method to recognise characters from an image are disclosed. The method includes receiving the at least one image, pre-processing the at least one image, extracting a plurality of characters from the corresponding at least one image, extracting at least one structure from the corresponding at least one image upon applying an edge detection technique to extract a structure, identifying a template based on extracted structure, subjecting the plurality of characters into a plurality of ensemble AI models to extract one of a plurality of texts, a plurality of non-textual data and a combination thereof, comparing a plurality of extracted plurality of texts, a plurality of non-textual data, or a combination thereof from the corresponding plurality of ensemble AI models with each other, generating a confidence score and validating one of the plurality of accurate texts, the plurality of accurate non-textual data, or a combination thereof.
1. A system to recognise one or more characters from at least one image comprising:
one or more processors;
an image receiving module operable by the one or more processors, and configured to receive the at least one image representative of a document;
a pre-processing module operable by the one or more processors, and configured to pre-process the at least one image to reduce noise and enhance one or more parameters associated with image quality;
a feature extraction module operable by the one or more processors, and configured to:
extract a plurality of characters from the corresponding at least one image using a character recognition technique;
extract at least one structure from the corresponding at least one image upon applying an edge detection technique to extract an X co-ordinate and a Y co-ordinate associated to the corresponding at least one structure, wherein the at least one structure is a representative of the document; and
identify a template based on at least one extracted structure upon comparing the template with a plurality of pre-stored templates stored in a database;
a character accuracy detection module operable by the one or more processors, and configured to:
subject the plurality of characters into a plurality of ensemble artificial intelligence (AI) models to extract one of a plurality of texts, a plurality of non-textual data and a combination thereof, from a plurality of extracted characters from each of the plurality of ensemble artificial intelligence (AI) models;
compare a plurality of extracted texts, a plurality of extracted non-textual data, or a combination thereof, from the corresponding plurality of ensemble artificial intelligence (AI) models with each other to generate a plurality of accurate texts, a plurality of accurate non-textual data, or a combination thereof; and
generate a confidence score for each of the plurality of accurate texts, each of the plurality of accurate non-textual data or a combination thereof; and
a character validation module operable by the one or more processors, and configured to validate one of the plurality of accurate texts, the plurality of accurate non-textual data, or a combination thereof, generated upon comparison based on the confidence score upon receiving insights by a user to recognise the one or more characters from the corresponding at least one image.
2. The system as claimed in claim 1 , wherein the parameters associated to image quality comprises to one of de-skewing of the at least one image, parameters associated to gamma value of the at least one image, contract of the at least one image, enhancement of the at least one image, or a combination thereof.
3. The system as claimed in claim 1 , wherein the plurality of characters comprises one of a text, numeric characters, non-numeric characters, special characters or a combination thereof.
4. The system as claimed in claim 1 , wherein the feature extraction module is configured to store the template in the database, if the identified template is not present in the database.
5. The system as claimed in claim 1 , comprising a representation module operable by the one or more processors, and configured to represent one of a plurality of validated texts, a plurality of validated non-textual data or a combination thereof in a required format.
6. A method for recognising one or more characters from at least one image comprising:
receiving, by an image receiving module, the at least one image representative of a document;
pre-processing, by a pre-processing module, the at least one image for reducing noise and enhancing one or more parameters associated with image quality;
extracting, by a feature extraction module, a plurality of characters from the corresponding at least one image using a character recognition technique;
extracting, by the feature extraction module, at least one structure from the corresponding at least one image upon applying an edge detection technique for extracting an X co-ordinate and a Y co-ordinate associated to the corresponding at least one structure;
identifying, by the feature extraction module, a template based on at least one extracted structure upon comparing the template with a plurality of pre-stored templates stored in a database;
subjecting, by a character accuracy detection module, the plurality of characters into a plurality of ensemble artificial intelligence (AI) models to extract one of a plurality of texts, a plurality of non-textual data and a combination thereof, from a plurality of extracted characters from each of the plurality of ensemble artificial intelligence (AI) models;
comparing, by the character accuracy detection module, a plurality of extracted plurality of texts, a plurality of non-textual data, or a combination thereof, from the corresponding plurality of ensemble artificial intelligence (AI) models with each other for generating a plurality of accurate texts, a plurality of accurate non-textual data, or a combination thereof;
generating, by the character accuracy detection module, a confidence score for each of the plurality of accurate texts, each of the plurality of accurate non-textual data or a combination thereof; and
validating, by a character validation module, one of the plurality of accurate texts, the plurality of accurate non-textual data, or a combination thereof, generated upon comparison based on the confidence score upon receiving insights by a user to recognise the one or more characters from the corresponding at least one image.
7. The method as claimed in claim 6 , wherein pre-processing the at least one image for reducing noise and enhance one or more parameters associated with image quality comprises pre-processing the at least one image for one of de-skewing of the at least one image, parameters associated to gamma value of the at least one image, contract of the at least one image, enhancement of the at least one image, or a combination thereof associated with the at least one image.
8. The method as claimed in claim 6 , wherein extracting the plurality of characters from the corresponding at least one image comprises extracting one of a text, numeric characters, non-numeric characters, special characters or a combination thereof from the corresponding at least one image.
9. The method as claimed in claim 6 , comprising storing, by the feature extraction module, the template in the database, if the identified template is not present in the database.
10. The method as claimed in claim 6 , comprising representing, by a representation module, one of a plurality of validated texts, a plurality of validated non-textual data or a combination thereof in a required format.