Method and apparatus for form identification and registration
View Patent ↗Aspects of the present invention relate to a machine learning system that is trained to identify forms, performing a method that includes receiving a form as an input image; identifying a field in the input image; identifying boundaries of the field; identifying locations of characters in the field; creating a two-dimensional space containing special characters; replacing the special characters with the characters in the field; identifying one or more keywords in the field based on identification of words and/or location of words; and responsive to an indication that the identifying one or more keywords yielded an incorrect result, updating the machine learning system. In another aspect, the machine learning system is used to identify forms, and can identify whether a form requires registration and, if registration is required, performing the registration.
1. A computer-implemented method of training a machine learning system to identify forms in document and form analysis, the method comprising:
a) receiving a form as an input image;
b) identifying a field in the input image;
c) identifying boundaries of the field;
d) identifying locations of characters in the field;
e) creating a two-dimensional space as a blown up representation of the field, the two-dimensional space containing special characters, the special characters being different from the characters in the field, the special characters being inserted in pixel locations so as to preserve the locations of the characters in the field;
f) replacing the special characters in the two-dimensional space with the characters in the field, the characters being positioned in locations in the two-dimensional space corresponding to the locations of the characters in the field, each character being related to a pixel location;
g) identifying one or more keywords in the field based on identification of words and/or location of words, wherein a) to g) enable form matching for the document and form analysis; and
h) responsive to an indication that the identifying one or more keywords yielded an incorrect result, updating the machine learning system.
2. The method of claim 1 , wherein updating the machine learning system comprises updating weights of nodes in the machine learning system.
3. The method of claim 1 , wherein the input image comprises a synthetic form or an image of a scanned form.
4. The method of claim 1 , wherein the identifying one or more keywords comprises reading the characters in the input image.
5. The method of claim 1 , wherein the identifying one or more keywords comprises using relative locations of characters in the 2D space to determine the one or more keywords.
6. The method of claim 1 , further comprising, responsive to a determination that there are more forms to be processed, receiving the next form and repeating the method.
7. A machine learning system to identify forms in document and form analysis, the machine learning system comprising at least one processor and a non-transitory memory that is programmed for the machine learning system to perform a method of document and form analysis comprising:
a) receiving a form as an input image;
b) identifying a field in the input image;
c) identifying boundaries of the field;
d) identifying locations of characters in the field;
e) creating a two-dimensional space as a blown up representation of the field, the two-dimensional space containing special characters, the special characters being different from the characters in the field, the special characters being inserted in pixel locations so as to preserve the locations of the characters in the field;
f) replacing the special characters in the two-dimensional space with the characters in the field, the characters being positioned in locations in the two-dimensional space corresponding to the locations of the characters in the field, each character being related to a pixel location;
g) identifying one or more keywords in the field based on identification of words and/or location of words, wherein a) to g) enable form matching for the document and form analysis; and
h) responsive to an indication that the identifying one or more keywords yielded an incorrect result, updating the machine learning system.
8. The system of claim 7 , wherein updating the machine learning system comprises updating weights of nodes in the machine learning system.
9. The system of claim 7 , wherein the input image comprises a synthetic form or an image of a scanned form.
10. The system of claim 7 , wherein the identifying one or more keywords comprises reading the characters in the input image.
11. The system of claim 7 , wherein the identifying one or more keywords comprises using relative locations of characters in the 2D space to determine the one or more keywords.
12. The system of claim 7 , further comprising:
responsive to the identifying the one or more keywords, determining a type of the form.
13. The system of claim 12 , further comprising, responsive to the responsive to the determining the type of form, determining whether the form requires registration.
14. The system of claim 7 , further comprising, responsive to a determination that there are more forms to be processed, receiving the next form and repeating the method.
15. The system of claim 13 , wherein, responsive to the determination that the form requires registration, performing registration on the form.