Systems and methods for automatically detecting a corner in a digitally captured image
AssigneeHewlett-Packard Development Company, L.P.
InventorTong Zhang
Life of this patent
Estimating the term…
Abstract
In one embodiment, the present invention is related to a method for processing a digitally captured image that comprises an imaged document. The method may comprise: transforming the digitally captured image into a binary image; searching the binary image to detect a plurality of edges of the imaged document; and analyzing the detected plurality of edges to determine at least one corner associated with the imaged document; wherein the transforming, searching, and analyzing are performed by programmable logic associated with a processor-based system.
Claims18
A method for processing a digitally captured image that comprises an imaged document, said method comprising:
- transforming said digitally captured image into a binary image, wherein a pixel of said binary image equals:
- (1) a first logical value when a corresponding pixel in said digitally captured image is associated with a value greater than a luminance threshold;
- or (2) a second logical value otherwise;
- searching said binary image to detect a plurality of edges of said imaged document;
- and
- analyzing said detected plurality of edges to determine at least one corner associated with said imaged document;
- wherein said transforming, searching, and analyzing are performed by programmable logic associated with a processor-based system.
A system for processing a digitally captured image that comprises an imaged document, said system comprising:
- means for transforming said digitally captured image into a binary image, wherein said means for transforming is operable to assign a first logical value to a pixel of said binary image that corresponds to a pixel of said digitally captured image that is associated with a value greater than a luminance threshold, and is operable to assign a second logical value otherwise;
- means for detecting edges of said imaged document from said binary image;
- and
- means for estimating at least one corner location of said imaged document from said detected edges.
A computer-readable medium comprising executable instructions for processing a digitally captured image that comprises an imaged document, said computer-readable medium comprising:
- code for applying a luminance threshold to said digitally captured image to construct a binary image, wherein a pixel of said binary image equals:
- (1) a first logical value when a corresponding pixel in said digitally captured image is associated with a value greater than said luminance threshold:
- or (2) a second logical value otherwise;
- code for detecting edges of said imaged document from said binary image, wherein said code for detecting is operable to search from each respective margin of said binary image for a change in value in said binary image to detect said edges;
- and
- code for determining at least one corner location from said detected edges.
Dependencies are inferred from the claim text.