System for extracting information from an identity card

• US 7,911,655 B2
• Filed: 10/06/2004
• Issued: 03/22/2011
• Est. Priority Date: 10/06/2004
• Status: Active Grant

• Current Assignee: Acuant Inc. (GB Group Plc)
• Original Assignee: CSSN Incorporated (GB Group Plc)
• Inventors: Hatzav, Iuval, Ben-Shahar, Ilan, Hatzav, Daniel, Sharaby, Giyora
• Primary Examiner(s): Worku; Negussie
• Application Number:
  US10/959,261
• Publication Number:
  US 20060072822A1
• Time in Patent Office: 2,358 Days
• Field of Search: 358/474, 358/473, 358/482, 358/483, 358/475, 358/509, 358/514, 358/518, 358/505, 358/520, 358/909.1, 358/909.6, 348/156, 348/96, 348/150, 348/207.1, 348/207.99, 348/211.14, 382/116, 382/117, 382/118, 382/284, 382/159, 382/190, 382/229, 382/312, 235/468, 235/487
• US Class Current: 358/474
• CPC Class Codes: G03B 15/00 Special procedures for taking photographs; Apparatus therefor
• CPC Class Codes: G03B 27/323 Copying cameras G03B27/34 - G03B27/70 take precedence

Abstract

The present invention provides a unique and novel system for acquiring digital images of identification documents such as drivers license, passports, or medical insurance records using a digital camera so as to establish a high resolution image and extracting data automatically with machine vision tools so as to acquire accurate data. The present invention teaches also a system that acquires multi-spectral images of both sides of the identification document.

Claims

1. A system for extracting information from an identity document comprising at least one light source for illuminating the identity document;

at least one digital camera comprising a lens and a two-dimensional sensor array for acquiring at least one two-dimensional image of the identity document; and

a processor for processing said at least one two-dimensional image, said processor configured to run an optical character recognition (OCR) routine for extracting textual information from the identity document in the form of alphanumerical information, wherein said extracting textual information comprises the following steps;

a) determining boundaries between text and background based on minimum contrast between the character and other colored data on the identity document;

b) extracting black and white glyphs using at least a portion of said determined boundaries;

c) applying said OCR routine on said black and white glyphs, thereby recognizing symbols and characters; and

d) sending said recognized symbols and characters as alphanumerical information.

2. A system according to claim 1, further comprising an indicator light for indicting the status of the system.

3. A system according to claim 1, further comprising a box for housing said digital camera and said light source;

and a communication link for connecting said box to said processor.

4. A system according to claim 1, acquiring at least red green and blue images of the identity document.

5. A system according to claim 4, further acquiring at least one image selected from the group of IR reflection image, UV reflection image and fluorescence image.

6. A system according to claim 1, further comprising mirrors for simultaneously imaging both sides of said document.

7. A system according to claim 1, wherein said at least one light source comprises at least two light sources for respectively illuminating both sides of said document;

and wherein said at least one digital camera comprises two digital cameras for respectively imaging both sides of the identity document.

8. A system according to claim 1, wherein said light source includes a light diffuser for even illumination of the identity document.

9. A system according to claim 1, wherein said light diffuser is a light reflector.

10. A system according to claim 9, including light baffle for preventing direct illuminating the identity document by said light source.

11. A system according to claim 1, wherein said extraction of textual data from the identity document is assisted by prior knowledge of the type of said document.

12. A system according to claim 1, wherein said acquisition of said at least one two-dimensional image is automatically initiated by sensing proper placement of the identity document.

13. A system according to claim 12, wherein said sensing proper placement of the identity document is performed using a sensor.

14. A system according to claim 12, wherein said sensing proper placement of the identity document is performed by acquiring images and performing at least partial image processing of said images.

15. A system according to claim 14, wherein said images used for sensing proper placement of identity card are of lower quality than images used for extracting information.

16. A system according to claim 1, wherein said alphanumerical information is used to save a user from manually inputting said information.

17. A system according to claim 1, wherein said extracted information is used to augment a database.

18. A system according to claim 1, wherein said extracted information is used as key words to perform a search.

19. A system according to claim 1, wherein said processor is configured the run pre-processing steps to facilitated better quality images of the identity document, before applying said OCR routine for extracting textual information from the identity document in the form of alphanumerical information.

20. A method for extracting information from an identity document, comprising the steps of:

a) providing a system for extracting information from an identity document including;

i. at least one light source for illuminating the identity document;

ii. at least one digital camera having a lens and a two-dimensional sensor array for acquiring at least one two-dimensional image of the identity document; and

iii. a processor for processing said at least one two-dimensional image;

b) determining boundaries between text and background based on minimum contrast between the character and other colored data on said image of the identity document;

c) extracting black and white glyphs using at least a portion of said determined boundaries;

d) applying an optical character recognition (OCR) routine on said black and white glyphs, thereby recognizing symbols and characters; and

e) sending said recognized symbols and characters as alphanumerical information;

wherein said determining of the type of the identity document and said identifying said textual field regions in the identity document, are performed before applying said OCR routine on said black and white glyphs.

21. A method according to claim 20, further comprising pre-processing steps to facilitated better quality images of the identity document, performed before applying said OCR routine on said black and white glyphs.

22. A method according to claim 21, wherein said pre-processing steps comprise at least one of the following steps:

i) determining the type of the identity document and removing the background from the image of the identity document;

ii) Calibrating the illumination parameters, before acquiring an image of the identity document for extracting information from the identity document; and

iii) correcting image distortions caused by know lens distortions.