Ambiguity reduction for image alignment applications

• US 9,430,457 B2
• Filed: 12/24/2014
• Issued: 08/30/2016
• Est. Priority Date: 12/24/2014
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Maltz, Martin S.
• Primary Examiner(s): NGUYEN, NGON BINH
• Application Number:
  US14/582,321
• Publication Number:
  US 20160188559A1
• Time in Patent Office: 615 Days
• Field of Search: 358/1.15
• US Class Current: 1/1
• CPC Class Codes: G06F 40/174 Form filling; Merging
• CPC Class Codes: G06F 40/186 Templates
• CPC Class Codes: H04N 1/3873 defined only by a limited number of coordinate points or parameters, e.g. corners, centre; for trimming
• CPC Class Codes: H04N 2201/0081 Image reader H04N2201/0091 - H04N2201/0094 take precedence

Abstract

According to exemplary systems and methods, a template is partitioned into blocks using an image processor. An image is scanned using an optical scanner. The image is aligned to the template. Image features are matched with template features in the blocks of the template. Displacement vectors are identified for differences of the image features from the template features. Normalized cross correlation (NCC) is determined between blocks of the image and each block of the template using the image processor. Peaks in the NCC are identified. A displacement vector is selected for a peak with highest NCC for each the block of the template. Ambiguous template features are identified in the image based on the displacement vector. Blocks are iteratively combined and NCC determined to remove the ambiguous template features.

Claims

1. A method comprising:

partitioning a template into blocks using an image processor, said template comprising a user-fillable pre-printed form;

scanning an image using an optical scanner, said image comprising said pre-printed form having user markings;

aligning said image to said template using said image processor;

matching image features with template features in said blocks of said template using said image processor;

identifying displacement vectors for differences of said image features from said template features using said image processor;

determining normalized cross correlation (NCC) between blocks of said image and each block of said template using said image processor;

identifying peaks in said NCC using said image processor;

selecting a displacement vector for a peak with highest NCC for each said block of said template using said image processor;

identifying ambiguous template features in said image based on said displacement vector using said image processor;

removing said ambiguous template features from said image using said image processor; and

iteratively combining blocks and determining NCC for each said block of said template using said image processor to remove said ambiguous template features from said image using said image processor.

2. The method according to claim 1, further comprising applying histogram equalization between said template and said image.

3. The method according to claim 1, further comprising creating an aligned image being the same size as said template using said image processor.

4. The method according to claim 3, further comprising:

removing features of said template that are common with features of said aligned image using said image processor; and

extracting remaining features in said aligned image using said image processor.

5. The method according to claim 4, said removing features of said template that are common with features of said aligned image further comprising removing only patterned lines from said image and leaving said user markings using said image processor, said remaining features comprising said user markings.

6. The method according to claim 1, said scanning said image using said optical scanner comprising obtaining a digital photograph of said image.

7. A method, comprising:

receiving a template into an image processor, said image processor comprising a special purpose machine that is specialized for processing image data, said template comprising a user-fillable pre-printed form;

providing electronic instructions to said image processor to cause said image processor to partition said template into blocks;

receiving a scan of an image into said image processor, said image comprising said pre-printed form having user markings;

providing electronic instructions to said image processor to cause said image processor to align said scan of said image to said template;

providing electronic instructions to said image processor to cause said image processor to match image features with template features in said blocks of said template;

providing electronic instructions to said image processor to cause said image processor to identify displacement vectors for differences of said image features from said template features;

providing electronic instructions to said image processor to cause said image processor to determine normalized cross correlation (NCC) between blocks of said image and each block of said template;

providing electronic instructions to said image processor to cause said image processor to identify peaks in said NCC;

providing electronic instructions to said image processor to cause said image processor to select a displacement vector for a peak with highest NCC for each said block of said template;

providing electronic instructions to said image processor to cause said image processor to identify ambiguous template features in said image based on said displacement vector;

providing electronic instructions to said image processor to cause said image processor to remove said ambiguous template features from said image; and

providing electronic instructions to said image processor to cause said image processor to iteratively combine blocks and determine NCC for each said block of said template to remove said ambiguous template features from said image.

8. The method according to claim 7, further comprising providing electronic instructions to said image processor to cause said image processor to apply histogram equalization between said template and said scan of said image.

9. The method according to claim 7, further comprising providing electronic instructions to said image processor to cause said image processor to create an aligned image, said aligned image being the same size as said template.

10. The method according to claim 9, further comprising:

providing electronic instructions to said image processor to cause said image processor to remove features of said template that are common with features of said aligned image; and

providing electronic instructions to said image processor to cause said image processor to extract remaining features in said aligned image.

11. The method according to claim 10, providing electronic instructions to said image processor to cause said image processor to remove features of said template that are common with features of said aligned image removing only patterned lines from said scan of said image and leaving said user markings, said remaining features comprising said user markings.

12. The method according to claim 7, said receiving a scan of said image into said image processor comprising obtaining a digital photograph of said image.

13. A system, comprising:

an image processor receiving a template and partitioning said template into blocks, said template comprising a user-fillable pre-printed form; and

an image capture device operatively connected to said image processor, said image capture device scanning an image to produce a scan, said image comprising said pre-printed form having user markings,

said image processor aligning said scan to said template,

said image processor matching features of said scan with features of said template in said blocks of said template,

said image processor identifying displacement vectors for differences of said features of said scan from said features of said template,

said image processor determining normalized cross correlation (NCC) between blocks of said scan and each block of said template,

said image processor identifying peaks in said NCC,

said image processor selecting a displacement vector for a peak with highest NCC for each said block of said template,

said image processor identifying ambiguous template features in said scan based on said displacement vector,

said image processor removing said ambiguous template features from said scan; and

said image processor iteratively combining blocks and determining NCC for each said block of said template to remove said ambiguous template features from said scan.

14. The system according to claim 13, said image processor applying histogram equalization between said template and said scan.

15. The system according to claim 13, further comprising a printing device operatively connected to said image processor, said printing device printing a form-based print job having user-fillable areas comprising patterned lines to print said user-fillable pre-printed form.

16. The system according to claim 13, said image capture device comprising one of scanners, cameras, photography equipment, facsimile machines, photo reproduction equipment, digital printing presses, and xerographic devices.

17. The system according to claim 13, further comprising:

said image processor creating an aligned image from said scan, said aligned image being the same size as said template

said image processor removing features of said template that are common with features of said aligned image; and

said image processor extracting remaining features in said aligned image.

18. The system according to claim 17, said image processor removing only patterned lines from said aligned image and leaving said user markings, said remaining features comprising said user markings.