zbar-lite

This module is used to provide an easy way to pack zbar python binding to wheel. And also provide some pythonic apis for users to use zbar easily.

Only supports image related functions.

For now , it is only designed for Python.

How to build and install

1. get zbar

git clone https://github.com/mchehab/zbar.git

It is recommended for you to run cmd below first to get the right config.h.

cd zbar
autoreconf -vfi
./configure --without-java --without-gtk --without-qt  --without-imagemagick  --disable-video --without-python

2. copy the source file we need

sh preparation.sh

There are two template config.h in the zbar_lite/config_template .

if you did not generate a config.h, we will copy one of them to ./src according to your OS when setup.

Before build, you should make sure that your gcc compiler is fit with your OS.

you can install gcc build env from https://sourceforge.net/projects/mingw-w64/files/ for windows

Notice that mingw64 and mingw-w64 is not the same thing.

It is recommended for you to install setuptools to install and build.

pip install setuptools wheel

Windows

I select x86_64-posix-seh-rev0 to build my wheel on Windows.

python setup.py build -c mingw32

to build whl

python setup.py build -c mingw32 bdist_wheel

to install

python setup.py build -c mingw32 install

Linux

to build whl

CC="gcc -std=gnu99" python setup.py bdist_wheel

to install

CC="gcc -std=gnu99" python setup.py install

Some errors you could meet:

1. Cannot find -lmsvcr140

if you build this whl in Windows with python setup.py build_ext --compiler=mingw32, you may meet an error that cannot find -lmsvcr140, as you can see in https://stackoverflow.com/questions/43873604/where-is-msvcr140-dll-does-it-exist.

I fixed it in the setup.py

2. Should be build by std99 or gnu99

CC="gcc -std=gnu99" python setup.py bdist_wheel

3. Do not support inverted Code

When the background is darker than the QR Code's foreground, it's called an inverted Code. These types of Codes typically have a dark background such as black, navy or dark grey. While a few scanners can read an inverted Code, some apps are not able to scan them including us.

4. Can not detect the barcode

You can do some preprocess before decoding it.
It is recommended for you to try the ways below:

1. Turn it to gray
2. Split the color channel, such as use the b or g or r channel separately

How to use

We provide several versions of whl right now. You can try to install via pip install zbar-lite.

example1

import zbar
import cv2

img_path='./test.jpg'

# create a reader
scanner = zbar.ImageScanner()

# configure the reader
scanner.parse_config('enable')

# obtain image data
pil = cv2.imread(img_path,cv2.IMREAD_GRAYSCALE)
height, width = pil.shape[:2]
raw = pil.tobytes()

# wrap image data
image = zbar.Image(width, height, 'Y800', raw)

# scan the image for barcodes
scanner.scan(image)

# extract results
for symbol in image:
    # do something useful with results
    print('decoded', symbol.type, 'text', '"%s"' % symbol.data)
    print('type {} text {} location {} quality {}'.format( symbol.type, symbol.data,symbol.location,symbol.quality))

# clean up
del(image)

example2

from zbar_helper.utils import decode, show_info
import cv2
image_path = "test.png"
img = cv2.imread(image_path)
print(decode(cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)))
show_info(decode(cv2.imread(image_path, cv2.IMREAD_GRAYSCALE)), img)

For more documents you can visit https://github.com/mchehab/zbar/tree/master/python

Some tests for more Barcode Tools https://github.com/xulihang/Barcode-Reading-Performance-Test