Visual bar code recognition method

• US 6,366,696 B1
• Filed: 12/20/1996
• Issued: 04/02/2002
• Est. Priority Date: 12/20/1996
• Status: Expired due to Term

• Current Assignee: NCR Corporation (NCR Voyix Corporation)
• Original Assignee: NCR Corporation (NCR Voyix Corporation)
• Inventors: Ming, John C., Su, Kevin K., Hertz, Lois H.
• Primary Examiner(s): COUSO, YON JUNG
• Application Number:
  US08/777,672
• Time in Patent Office: 1,929 Days
• Field of Search: 382/182, 382/183, 382/309, 382/181, 382/295, 382/312, 235/462, 235/470
• US Class Current: 382/183
• CPC Class Codes: G06V 30/10 Character recognition
• CPC Class Codes: G06V 30/146 Aligning or centring of the image pick-up or image-field

Abstract

A visual bar code recognition method which combines conventional decoding techniques with optical character recognition (OCR). The visual bar code recognition method captures an image of an object containing a bar code. Regardless of the orientation of the bar code within the field-of-view, the system detects the presence of the bar code, and decodes it using the bar/space patterns. It then produces an independent decoding of the human-readable numbers printed on the bar code using OCR. From these two decodings, it determines the identity of the object. It verifies this identity by comparing the physical characteristics of the object from the image with the known features of the product.

Claims

1. A method of decoding a bar code to identify an item having the bar code on it comprising the steps of:

capturing an image of the item;

locating the bar code label in an area of the image;

decoding the bar code label located in the area of the image to produce a first set of characters;

performing optical character recognition of the area of the image to produce a second set of characters; and

comparing the first set of characters to the second set of characters to identify the item.

2. The method as recited in claim 1, further comprising the steps of:

determining predetermined features of the item from the image; and

comparing the features to features stored within a database to verify the identity of the item.

4. The method of claim 1 wherein the step of capturing an image of the item is performed utilizing a camera to scan a substantial portion of the item including the bar code.

5. The method of claim 1 wherein the step of locating the bar code label further comprises the steps of determining the location of a dense concentration of high-strength edges within the image and analyzing whether said edges have a similar phase.

6. The method of claim 1 wherein the bar code has an associated human readable character string located in close proximity to a series of bars and spaces defining a coded portion of the bar code, said step of decoding the bar code label comprises a decoding of the series of bars and spaces, and said step of performing optical character recognition comprises a recognition of said associated human readable character string.

7. The method of claim 6 wherein identification of the item occurs only if the first set of characters and the second set of characters sufficiently match.

8. The method of claim 2 wherein said predetermined features correspond to physical characteristics of the item.

9. The method of claim 8 wherein identification of the item occurs only if the first set of characters and the second set of characters sufficiently match and said predetermined features sufficiently match the features stored within the database.

3. A method of decoding a bar code on an item comprising the steps of:

providing a gray-scale image of the item;

determining gray levels of pixels within the image;

filtering out pixels having gray levels below a predetermined threshold gray level;

determining an area of the image having a density of pixels with gray levels above the predetermined threshold density to locate the bar code;

determining an orientation of the area;

tracing a plurality of gray-scale lines through the area;

transforming the image into a binary image;

transforming the gray-scale lines to binary lines;

decoding the binary lines to produce a first set of characters;

performing optical character recognition of the area to produce a second set of characters;

comparing the first set of characters to the second set of characters to identify the item;

determining predetermined features of the item from the image; and

comparing the features to features stored within a database to verify the identity of the item.

10. A visual recognition apparatus for decoding a bar code to identify an item, the apparatus comprising:

a camera for capturing an image of the item;

a bar code label locator for locating an area within the image including the bar code;

a decoder for decoding the bar code and to produce a first set of characters corresponding to the bar code;

an optical character recognizer for recognizing a human-readable string of characters associated with the bar code and also located in said area, and for producing a second set of characters corresponding to the human-readable string of characters; and

a processor for comparing the first set of characters and the second set of characters and to identify the item if the first set of characters and the second set of characters sufficiently match.

11. The apparatus of claim 10 wherein the processor is further operable to determine predetermined features of the item from the image, and to compare the predetermined features to features stored within a database to further verify the identity of the item.

12. The apparatus of claim 10 wherein the bar code label locator further comprises an edge filter for locating an area within the image having a dense concentration of high-strength edges.

13. The apparatus of claim 12 further comprising an analyzer for analyzing whether said edges have a similar phase.

14. The apparatus of claim 11 wherein the predetermined features correspond to physical features of the item.