US 11,776,248 B2
Systems and methods for automated document image orientation correction
Rahul Bhaskar, Rancho Santa Margarita, CA (US); Daryl Seiichi Furuyama, Irvine, CA (US); and Daniel William James, Costa Mesa, CA (US)
Assigned to Optum, Inc., Minnetonka, MN (US)
Filed by Optum, Inc., Minnetonka, MN (US)
Filed on Oct. 17, 2022, as Appl. No. 18/47,045.
Application 18/047,045 is a continuation of application No. 16/935,880, filed on Jul. 22, 2020, granted, now 11,495,014.
Prior Publication US 2023/0135212 A1, May 4, 2023
Int. Cl. G06V 10/98 (2022.01); G06N 20/00 (2019.01); G06V 30/40 (2022.01); G06V 10/24 (2022.01); G06F 18/21 (2023.01); G06F 18/28 (2023.01); G06V 30/10 (2022.01)
CPC G06V 10/98 (2022.01) [G06F 18/21 (2023.01); G06F 18/217 (2023.01); G06F 18/28 (2023.01); G06N 20/00 (2019.01); G06V 10/242 (2022.01); G06V 30/40 (2022.01); G06V 30/10 (2022.01); G06V 2201/10 (2022.01)] 18 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
providing, by one or more processors, a first rotated machine readable text data object of a plurality of rotated machine readable text data objects to a natural language processing (NLP) engine, wherein the first rotated machine readable text data object is generated by:
(a) generating, by applying an optical character recognition (OCR) process, initial machine readable text for an original image data object,
(b) generating, using one or more machine learning models, an initial quality score for the initial machine readable text, wherein the initial quality score indicates a probability that an error in the initial machine readable text is attributable to an image orientation associated with the original image data object,
(c) responsive to a determination that the initial quality score does not satisfy one or more quality criteria, generating a plurality of rotated image data objects, wherein (i) each of the plurality of rotated image data objects corresponds to a different rotational position and (ii) each of the plurality of rotated image data objects comprises the original image data object rotated to a corresponding rotational position,
(d) generating the plurality of rotated machine readable text data objects for the plurality of rotated image data objects,
(e) generating, using one or more machine learning models, a plurality of rotated quality scores comprising a rotated quality score for each of the plurality of rotated machine readable text data objects, and
(f) determining that a first rotated quality score of the plurality of rotated quality scores satisfies the one or more quality criteria, wherein (i) the first rotated quality score corresponds to the first rotated machine readable text data object and (ii) determining that the first rotated quality score satisfies the one or more quality criteria indicates that the first rotated machine readable text data object is to be provided to the NLP engine.