Method and Apparatus for Correcting Decoding Errors in Machine-Readable Symbols
AssigneeSeiko Epson Corporation
InventorsJia Li, Mikhail Brusnitsyn, Eunice Poon, Seungsik Kwak
Life of this patent
Estimating the term…
Abstract
The present invention is a system and method for decoding an image of a bar code. Decoding the barcode includes tokenizing a plurality of pixels in the image of the barcode based upon a plurality of thresholds to form a first set of tokens. Decoding the barcode also includes re-tokenizing the plurality of pixels in the image of the barcode based upon the intensity of the pixels in the plurality of tokens, and the relativity intensity of neighboring tokens, to form a second set of tokens.
Claims16
A method of decoding an image of a barcode comprising the following steps:
- (a) tokenizing a plurality of pixels in the image of the barcode based upon at least an initial fixed set of intensity thresholds to form a first set of tokens, said first set of tokens including white tokens identifying decoded white spaces in the image and black tokens identifying decoded black bars in the image;
- (b) re-tokenizing the plurality of pixels in the image of the barcode based upon a first new set of thresholds determined from the intensity of pixels in the first set of tokens, and the relativity intensity of one of its neighboring re-tokenized tokens, to form a second set of tokens.
A method for decoding a barcode comprising:
- identifying a barcode candidate within an image consisting of an array of pixels;
- tokenizing the array of pixels into a plurality of initial black tokens and a plurality of initial white tokens using a plurality of initial thresholds, wherein a plurality of initial tokens consists of the initial black tokens and the initial white tokens;
- identifying an intensity of the darkest pixel among the brightest pixel in each of the plurality of initial white tokens as a first tokenization parameter, Tw;
- identifying an intensity of brightest pixel among the darkest pixel in each of the plurality of initial black tokens as a second tokenization parameter, Tb;
- calculating a tokenization contrast parameter, C, that is proportional to two-thirds of the difference between the first tokenization parameter Tw and the second tokenization parameter Tb;
- calculating a third tokenization parameter, t1, as equal to a first fixed amount higher than the second tokenization parameter, Tb;
- calculating a fourth tokenization parameter, t2, as equal to a second fixed amount lower than the first tokenization parameter Tw;
- calculating a fifth tokenization parameter, t2′, as equal to the sum of 0.75 times the first tokenization parameter Tw and 0.25 times the second tokenization parameter Tb;
- calculating a sixth tokenization parameter t1′ as equal to the sum of 0.25 times the first tokenization parameter Tw and 0.75 times the second tokenization parameter Tb;
- identifying a pixel as a black pixel if an intensity of the pixel is less than the third tokenization parameter t1 or if the intensity of the pixel is less than the sixth tokenization parameter t1′ and a difference between the intensity of the previous token and the intensity of the pixel is greater than the tokenization contrast parameter C;
- and
- identifying a pixel as a white pixel if an intensity of the pixel is greater than the fourth tokenization parameter t2 or if the intensity of the pixel is greater than the fifth tokenization parameter t2′ and difference between the intensity of the previous token and the intensity of the pixel is greater than the tokenization contrast parameter C.
A computer readable medium encoded with instructions for a processor to decode an image of a barcode comprising:
- instructions for tokenizing a plurality of pixels in the image of the barcode based upon a plurality of thresholds to form a first set of tokens;
- instructions for decoding the first set of tokens to form a first code;
- instructions for verifying that the first code is a valid barcode;
- instructions for re-tokenizing the plurality of pixels in the image of the barcode based upon the intensity of the pixels in the plurality of tokens, and the relativity intensity of neighboring tokens, to form a second set of tokens, if the first code is not a valid barcode.
Dependencies are inferred from the claim text.