This paper tells us about locating and decoding EAN-13 Barcodes from images captured
by digital cameras.
EAN-13 barcodes are a 1-D type of barcodes used for identification of books. This paper
thus aims to enable mobile phones with cameras to work as a barcode scanning device.
The phone, programmed to interpret barcode images, will generate the barcode value
which is used to identify the product. This paper contains a brief explanation of the EAN-
13 barcode (from the decoding point of view instead of encoding), proposed algorithm
for locating and decoding barcode imagesand the concluding remarks.
EAN is the acronym for European Article Numbering. This is used for identification
of products.This barcode has 13 numeric …show more content…
This barcode has 13 characters known as digits. Usually the first two digits are the
country code which are followed by 9 or 10 characters depending on the country code.
An example of an EAN-13 barcode is given below.
This EAN-13 barcode always begins and ends with a black-white-black (101) pattern
which are known as left hand guard and right hand guard. There is also a center guard
with pattern white-black-white-black-white (01010). These guards are slightly longer
lines compared to rest bars. All bars in this guard pattern are of width 1.
Each bar has a fixed width pattern given to it which is further used to calculate. Each
pattern consists of 4 bars, alternating with white and black.
The bar pattern for digits to the left hand of the center guard always start with a white
bar and end with black, whereas on the right of the center guard bar pattern starts with
a black bar and ends with a white bar. A bar pattern should always be of total width 7
which can represent a no. in binary.
The …show more content…
The algorithm is as follows
1. Preprocessing
2. Barcode locator
3. Barcode Decoder
Preprocessing includes checking if the image is in grayscale or not. If not, then luminance
component will be extracted.
The input is divided into blocks of equal size. Then blocks are created which are
then binarized using the Otsu thresh holding technique. Then an operation known
as skeletoning is done on each block of binary data. The orientation of each block is
calculated and compared in order to check the parallel property.
If all the angles in a block are of similar value we come to a conclusion that the block
contains a barcode. A group of blocks having similar parallel line patters indicate
presence of barcode. In this manner a barcode is located using the orientation angle of
each bar and the region is cropped and