Gap shifting for automatic recognition of tabular text
AssigneeAmazon Technologies, Inc.
InventorsWei You, Dirk Ryan Padfield, Gurumurthy Swaminathan
Life of this patent
Estimating the term…
Abstract
Disclosed are various embodiments for improving optical character recognition approaches through the use of gap shifting. A text detection process is performed upon an image to detect a first region of text. A second region that is in line with the first region is shifted to reduce a gap between the first region and the second region, thereby creating a modified image. The text detection process is performed upon the modified image in order to detect text within the second region.
Claims20
A non-transitory computer-readable medium embodying a program executable in at least one computing device that comprises a processor and a memory, wherein the program, when executed, causes the at least one computing device to at least:
- receive an image that includes text in a tabular form;
- perform an automated text detection process upon the image to detect a first region of text in a row;
- determine a second region in which the automated text detection process failed to detect text that is also in the row;
- shift the second region to reduce a whitespace gap between the first region and the second region, thereby creating a modified image;
- perform the automated text detection process upon the modified image to detect text in the second region;
- perform an automated text recognition process on the first region and the second region to determine a first character string from the first region and a second character string from the second region;
- and
- generate a representation of the text in the tabular form, wherein the first character string is visually separated from the second character string according to an amount of whitespace by which the second region was linearly shifted.
A system, comprising:
- at least one computing device comprising a processor and a memory;
- and
- at least one application executed in the at least one computing device, the at least one application causing the at least one computing device to at least:
- perform an automated text detection process upon an image to detect a first region of text;
- determine a second region in which the automated text detection process failed to detect text;
- shift the second region relative to the first region to reduce a gap between the first region and the second region, thereby creating a modified image;
- and
- perform the automated text detection process upon the modified image.
A method, comprising:
- receiving, by at least one computing device comprising a processor and a memory, an image that includes text;
- performing, by the at least one computing device, an automated text detection process upon the image to detect a first region of text;
- determining, by the at least one computing device, a second region in which the automated text detection process failed to detect text that is in line with the first region of text;
- shifting, by the at least one computing device, the second region to reduce a gap between the first region and the second region, thereby creating a modified image;
- and
- performing, by the at least one computing device, the automated text detection process upon the modified image.
Dependencies are inferred from the claim text.