Method and apparatus for barcode recognition in a digital image

• US 4,992,650 A
• Filed: 03/29/1990
• Issued: 02/12/1991
• Est. Priority Date: 03/29/1990
• Status: Expired due to Term

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Somerville, Peter J.
• Primary Examiner(s): NOT, DEFINED
• Assistant Examiner(s): NOT, DEFINED
• Application Number:
  US07/500,980
• Time in Patent Office: 320 Days
• Field of Search: 235/462, 235/470, 235/463, 235/472
• US Class Current: 235/462.1
• CPC Class Codes: G06K 7/1093 sensing, after transfer of the image of the data-field to an intermediate store, e.g. storage with cathode ray tube
• 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 recognition method and apparatus is disclosed which provides information on what type of barcode is to be recognized and the strategy to be applied in identifying and decoding the barcode. For N raster lines of a dimension "x", an array of one byte sums of dimension "x" is created. The sums are the sum of the "1" bits in the rectangular binary image input array in the N lines of the image. A new array can be generated which will have a dimension equal to the number of changes from black to white in the original "x-y" digital image. The values in this array will represent the length of each run of same color pels. These runs are then normalized according to rules of a particular barcode standard being handled.

Claims

1. A computer implemented method for recognizing a barcode contained in data representing a digital image, comprising the steps of:

buffering digital image data for N raster lines of a length of x pels, where N is an integer greater than 2 and x is an integer greater than 2;

forming an array of x sums from said N raster lines of buffered data;

comparing each sum with a threshold value and attributing it as representing a black area if greater than said threshold value or attributing it as representing a white area if less than said threshold value;

forming a second array of data representing said black and white regions;

normalizing said sequence of black and white regions to a barcode standard;

dividing said normalized regions by a value related to the total number of spatial elements for the barcode standard;

characterizing each spatial area as a long or short run;

decoding each barcode character and generating an alphanumeric string represented by said barcode.

2. A computer implemented method for recognizing a barcode contained in data representing a digital image, comprising the steps of:

inputting data representing a plurality of M raster lines of a digital image;

determining whether one raster line of said plurality contains a barcode;

converting said one raster line into a run-end array;

determining if there are N elements in said run-end array of said one raster line;

converting said plurality of M raster lines into a plurality of M run-end arrays, representing a plurality of pel columns of said digital image;

summing individual pel columns of said plurality of pel columns for said M raster lines;

attributing one pel column of said plurality of pel columns as black, if a sum of said one pel column is greater than an arithmetic function of M;

calculating run lengths for pel columns of said plurality of pel columns attributed as black;

grouping a plurality of P elements from said calculated run lengths;

identifying a barcode character from said P elements

3. The method of claim 2, wherein said step of determining whether one raster line of said plurality contains a barcode, comprises determining whether at least a predetermined number of run-ends occur in said one raster line.

4. The method of claim 2, wherein said step of inputting data further comprises, selecting the value of the number M raster lines, based upon the expected height of the barcode.

5. The method of claim 2, wherein said step of inputting data further comprises, selecting the value of the number M of raster lines, based upon the anticipated maximum level of skew of the barcode.

6. The method of claim 2, wherein said step of attributing one pel column as black, further comprises selecting said arithmetic function of M to be based on scanner contrast values for differentiating between black and white.

7. An computer implemented apparatus for recognizing a barcode contained in data representing a digital image, comprising:

a buffer for buffering digital image data for N raster lines of a length x, where N is an integer greater than 1 and x is an integer greater than 2;

means for forming an array of x sums from said N raster lines of buffered data;

a comparator for comparing each sum with a threshold value and attributing it as representing a black area if greater than said threshold value or attributing it as representing a white area if less than said threshold value;

means for forming a second array of data representing said black and white regions;

means for normalizing said sequence of black and white regions to a barcode standard;

dividing means for dividing said normalized regions by a value related to the total number of spatial elements for the barcode standard;

means for characterizing each spatial area as a long or short run;

decoder for decoding each barcode character and generating an alphanumeric string represented by said barcode.