Barcode location determination

• US 4,948,955 A
• Filed: 12/22/1988
• Issued: 08/14/1990
• Est. Priority Date: 12/22/1988
• Status: Expired due to Fees

• Current Assignee: The Boeing Co.
• Original Assignee: The Boeing Co.
• Inventors: Wang, Robert C., Lee, Shih-Jong J.
• Primary Examiner(s): Pitts, Harold I.
• Application Number:
  US07/289,662
• Time in Patent Office: 600 Days
• Field of Search: 235/462, 235/463
• US Class Current: 235/462.08
• CPC Class Codes: G06K 7/10851 Circuits for pulse shaping, amplifying, eliminating noise signals, checking the function of the sensing device basic electronic circuitry H03
• CPC Class Codes: G06K 7/14 using light without selection of wavelength, e.g. sensing reflected white light G06K7/10831 - G06K7/1097 take precedence

Abstract

A method of determining the location of a barcode block comprising the steps of reducing an input image, rejecting background clutter, detecting barcode regions, extracting barcode moment-based features, and determining four corner points of said barcode block. This method of determining the location of a barcode block can be implemented in a single PC card and achieve a throughput of more than 16 documents per second.

Claims

1. A method of determining the location of a barcode block comprising the steps of:

reducing input image data into a down-sampled image;

removing background clutter from said down-sampled image thereby creating a clutter-free, down-sampled image;

detecting barcode regions from said clutter-free, down-sampled image;

extracting connected, collinear components, along with their associated moment-based features, from said barcode regions; and

grouping said connected, collinear components into barcode blocks.

2. A method for determining the location of a barcode block as recited in claim 1 wherein said reducing step includes a decimation which skips a predetermined number of pixels and lines.

3. A method for determining the location of a barcode block as recited in claim 2 wherein said reducing step averages selected pixels and then thresholds to convert said down-sampled image into binary bit-map form.

4. A method for determining the location of a barcode block as recited in claim 1 wherein said input image data is operated on by said reducing step to reduce said input image data by a ratio of 4 to 1 in each dimension, thus resulting in an overall 16 to 1 reduction.

5. A method of determining the location of a barcode block as recited in claim 1 wherein said step of removing background clutter comprises a first operation to remove large background clutter, and a second operation to remove small background clutter.

6. A method of determining the location of a barcode block as recited in claim 5 wherein said first operation consists of a morphological opening residue operation by a large symmetric structuring element.

7. A method of determining the location of a barcode block as recited in claim 6 wherein said large symmetric structuring element is a disc with a radius of six picture elements.

8. A method of determining the location of a barcode block as recited in claim 5 wherein said second operation consists of a morphological opening residue operation by a small symmetric structuring element.

9. A method of determining the location of a barcode block as recited in claim 8 wherein said small symmetric structuring element is a disc with a radius of two picture elements.

10. A method of determining the location of a barcode block as recited in claim 1 wherein said step of detecting barcode regions comprises the following sub-steps:

performing a morphological dilation by a large structuring element to generate a barcode region-of-interest mask;

eliminating image features outside said barcode region-of-interest mark; and

performing a second morphological erosion by a small symmetric structuring element to detect said barcode regions.

11. A method of determining the location of a barcode block as recited in claim 10 wherein said large structuring element is a box having sides equal in length to fifteen picture elements.

12. A method of determining the location of a barcode block as recited in claim 10 wherein said small symmetric structuring element is a disk with a radius of one picture element.

13. A method of determining the location of a barcode block as recited in claim 10 wherein said sub-step of eliminating image features outside said barcode region-of-interest mask is performed by an AND gate having a first input supplied by said barcode region-of-interest mask, and a second input supplied by said clutter-free, down-sampled image.

14. A method of determining the location of a barcode block as recited in claim 1 wherein said associated moment-based features include centroid location, area, major axis length, and minor axis length.

15. A method of determining the location of a barcode block as recited in claim 1 wherein said grouping step comprises a unary test and a binary test.

16. A method of determining the location of a barcode block as recited in claim 15 wherein said unary test compares the heights of said connected, collinear components to a known height of a barcode and filters out those components which are either too large or too small as compared to said known height.

17. A method of determining the location of a barcode block as recited in claim 16 wherein said unary test determines the barcode axis of a component.

18. A method of determining the location of a barcode block as recited in claim 15 wherein said binary test determines the length of said barcode block.

19. A method of accurately detecting barcode locations in order to permit accurate barcode reading comprising the steps of:

receiving a digital image of a scanned document as an input;

reducing said input thereby producing a down-sampled image;

rejecting background clutter on said down-sampled image, thereby creating a clutter-free, down-sampled image;

detecting barcode regions from said clutter-free, down-sampled image;

applying a connected components feature extraction algorithm to detect connected components within said barcode regions; and

applying a barcode clustering algorithm which utilizes said connected components as inputs, and groups collinear components into barcode blocks.

20. A method of accurately detecting barcode locations as recited in claim 19 wherein said scanned document is a shipping label.

21. A method of accurately detecting barcode locations as recited in claim 19 wherein said accurate barcode reading is performed at a throughput of greater than 16 documents per second.

22. A method of locating a barcode from an input stage comprising the step of:

updating a hypothesized barcode block by merging connected components into said hypothesized barcode block.

23. A method of locating a barcode from an input stage as recited in claim 22 wherein a component is merged into said hypothesized barcode block only if:

(a) a difference between a barcode axis angle of said component and a barcode axis angle of a hypothesized barcode block is less than a first threshold value;

(b) a shortest distance between said component and said hypothesized barcode block is less than a second threshold value; and

(c) a perpendicular distance from a centroid of said component to a barcode axis of said hypothesized barcode block is less than a third threshold value.

24. A method of locating a barcode from an input stage as recited in claim 22 wherein the height of said hypothesized barcode block is predetermined.

25. A method of locating a barcode from an input stage as recited in claim 24 wherein the length of said hypothesized barcode block is updated to include all connected components.

26. A method of locating a barcode from an input stage as recited in claim 22 wherein the length of said hypothesized barcode block determines the length of said barcode.

27. A method of locating a barcode from an input stage as recited in claim 22 wherein when said updating step is completed, four corner points are determined for said hypothesized barcode block, and said four corner points represent a location boundary for said barcode.