Pure adversarial approach for identifying text content in images
AssigneeTrend Micro Incorporated
InventorJonathan James Oliver
Life of this patent
Estimating the term…
Abstract
A pure adversarial optical character recognition (OCR) approach in identifying text content in images. An image and a search term are input to a pure adversarial OCR module, which searches the image for presence of the search term. The image may be extracted from an email by an email processing engine. The OCR module may split the image into several character-blocks that each has a reasonable probability of containing a character (e.g., an ASCII character). The OCR module may form a sequence of blocks that represent a candidate match to the search term and calculate the similarity of the candidate sequence to the search term. The OCR module may be configured to output whether or not the search term is found in the image and, if applicable, the location of the search term in the image.
Claims17
A computer-implemented method of identifying text content in images, the method comprising:
- receiving an input image;
- splitting the image into a plurality of character blocks, each character block in the plurality of character blocks containing pixel information that may represent one or more characters;
- calculating a probability that a character block in the plurality of character blocks includes a character;
- forming a candidate sequence of character blocks from the plurality of character blocks, the candidate sequence of character blocks representing a candidate match for a search term;
- and
- comparing the candidate sequence of character blocks to the search term to determine if the search term is present in the candidate sequence of character blocks.
A computer having a memory and a processor configured to execute computer-readable program code in the memory, the memory storing:
- an email processing engine configured to receive an email and extract an image from the email;
- and
- a pure adversarial optical character recognition (OCR) module configured to receive a search term and the image and to search the image for the search term;
- wherein the OCR module is configured to split the image into a plurality of character blocks, calculate a probability that a character block in the plurality of character blocks includes a character, create a candidate sequence of character blocks from the plurality of character blocks to form a candidate match for the search term, and compare the candidate sequence of character blocks to the search term to determine if the search term is present in the candidate sequence of character blocks.
A computer-implemented method of identifying text content in images, the method comprising:
- extracting an image from an email;
- and
- searching the image for presence of a search term by splitting the image into a plurality of character blocks, each character block in the plurality of character blocks containing pixel information that may represent a character, calculating a probability that a character block in the plurality of character blocks includes a character, forming a candidate sequence of character blocks from the plurality of character blocks to form a candidate match for the search term, and comparing the candidate sequence of character blocks to the search term to determine if the search term is present in the candidate sequence of character blocks.
Dependencies are inferred from the claim text.