DATA INPUT FROM SERIES OF IMAGES APPLICABLE TO TEMPLATE DOCUMENT Russian patent published in 2017 - IPC G06K9/46 G06K9/68 G06T1/00 

FIELD: physics.

SUBSTANCE: group of inventions relates to optical character recognition (OCR) technologies. A method for retrieving data from a series of images of a template document comprises of the step of obtaining a current image from a series of images of a copy of the template document by the processing unit of the computer system, where the template document has at least one static element and at least one information field. Further, according to the method, the optical character recognition (OCR) of the current image is performed to obtain the recognized text and the corresponding coordinates of each character of the recognized text. And also, they determine the coordinate transformation parameters for converting the coordinates of the current image to the coordinates of the template based on the coordinates of the reference points of the current image and the template. The template contains the text and coordinates of at least one static element of the template document and coordinates of at least one information field of the template document.

EFFECT: improving the quality of extracted data and providing noise immunity.

27 cl, 6 dwg