Bar code reader

• US 4,988,852 A
• Filed: 03/19/1990
• Issued: 01/29/1991
• Est. Priority Date: 07/05/1988
• Status: Expired due to Fees

• Current Assignee: Intermec Corporation (Honeywell International Inc.)
• Original Assignee: Teknekron Transportation Systems Incorporated
• Inventors: Krishnan, Kalyan V.
• Primary Examiner(s): Pitts, Harold
• Application Number:
  US07/496,167
• Time in Patent Office: 316 Days
• Field of Search: 235/463, 235/462, 235/472, 235/494, 382/12, 382/48
• US Class Current: 235/462.08
• CPC Class Codes: G06K 7/0166 by means of clock-signals derived from the code marks, e.g. self-clocking code
• CPC Class Codes: G06K 7/14 using light without selection of wavelength, e.g. sensing reflected white light G06K7/10831 - G06K7/1097 take precedence
• CPC Class Codes: G06K 7/1443 locating of the code in an image
• CPC Class Codes: G06K 7/1456 determining the orientation of the optical code with respect to the reader and correcting therefore

Abstract

A system for finding and reading a bar code, arbitrarily located on a document, working from an electronic image of the document represented by a pixel array such as generated by a document scanner. A one-dimensional texture analysis is performed in the scanning direction by defining a pattern of test segments and searching for characteristic ranges of transistions within those test segments. If a horizontal test segment pattern fails to locate a bar code region, a test segment pattern skewed with respect to the horizontal is applied. If test patterns skewed with respect to the horizontal fail to locate a bar code region, then a state-transition analysis is applied in the direction perpendicular to the scan direction and in directions skewed with respect to that perpendicular to search for a terminal symbol characterizing the bar code.

Claims

1. A method of automatically detecting a bar code or the like on a document from an image of the document, said bar code being disposed at an unknown position on the document and said image being represented by a pixel array, comprising the steps of;

automatically locating a candidate bar code within a portion of said pixel array;

defining a baseline of said candidate bar code; and

decoding said bar code in the direction of said baseline.

2. The method of claim 1 wherein said locating step comprises the steps of:

defining a pattern of segments within said pixel array;

counting the number of transitions of said pixel array within each of said segments of said pattern; and

comparing said numbers with predetermined ranges of transition numbers characteristic of bar code, whereby said pattern of segments overlies a candidate bar code if the corresponding numbers of transitions fall within said predetermined characteristic ranges.

3. The method of claim 2 wherein said pattern comprises a first linear segment and a plurality of subsegments thereof.

4. The method of claim 2 wherein said pattern of segments is defined only in selected rows of said array and only in selected portions of said selected rows, whereby said locating step is performed only in said selected portions and not performed in the non-selected rows and portions, so as to reduce the time to locate said candidate bar code within said pixel array.

5. The method of claim 2, wherein said pixel array is addressed in a predetermined direction, and said linear segments are composed of co-linear pixels in said predetermined direction.

6. The method of claim 5 further comprising the steps of:

determining a pattern of segments skewed with respect to said predetermined direction; and

performing said counting and comparing steps with respect to said skewed pattern, if said pattern of segments in said predetermined direction fails to locate a candidate bar code.

7. The method of claim 6, wherein said skewed pattern of segments is composed of blocks of pixels offset from one another perpendicular to said predetermined direction.

8. The method of claim 6, wherein said bar code includes a characteristic terminal symbol, further comprising the step of:

when said counting and comparing steps with respect to said skewed pattern fail to locate a candidate bar code, subjecting successions of pixels from said pixel array in a direction generally perpendicular to said predetermined direction to state-transition analysis wherein said successions of pixels are compared with successions of states characteristic of said terminal symbol.

9. The method of claim 8, further comprising the step of:

repeating said state-transition analysis with a succession of pixels skewed with respect to said perpendicular direction when said state-transition analysis along said generally perpendicular direction fails to locate a candidate bar code.

10. The method of claim 1 wherein said bar code includes a characteristic terminal symbol and said locating step comprises the step of:

subjecting a succession of pixels from said pixel array to state-transition analysis wherein said succession of pixels is compared with a succession of states characteristic of said terminal symbol.

11. A method of systematically detecting a bar code or the like on a document from an image of the document, said image being represented by a pixel array addressed in a predetermined addressing direction and said bar code being disposed at an unknown position on said document and in an unknown orientation within a predetermined angular range with respect to said addressing direction, said method comprising the steps of:

defining a first pattern of segments within said pixel array extending substantially in said predetermined addressing direction;

counting the number of transitions of said pixel array within each of said segments of said pattern;

comparing said numbers with predetermined ranges of transition numbers characteristic of bar codes, whereby said first pattern of segments overlies a candidate bar code if the corresponding numbers of transitions fall within said predetermined characteristic ranges;

if said first pattern of segments fails to locate a candidate bar code, defining a second pattern of segments skewed with respect to said predetermined direction at an angle within said predetermined angular range;

performing said counting and comparing steps with respect to said second pattern and

if said second pattern of segments fails to locate a candidate bar code, defining further patterns of segments at successively greater angles of skew within said predetermined angular range, and performing said counting and comparing steps with respect to said further patterns of segments so as to locate a candidate bar code.