How to decode a QR-code image in (preferably pure) Python?

Question

TL;DR: I need a way to decode a QR-code from an image file using (preferable pure) Python.

I've got a jpg file with a QR-code which I want to decode using Python. I've found a couple libraries which claim to do this:

PyQRCode (website here) which supposedly can decode qr codes from images by simply providing a path like this:

import sys, qrcode
d = qrcode.Decoder()
if d.decode('out.png'):
    print 'result: ' + d.result
else:
    print 'error: ' + d.error

So I simply installed it using sudo pip install pyqrcode. The thing I find strange about the example code above however, is that it only imports qrcode (and not pyqrcode though) Since I think qrcode refers to this library which can only generate qr-code images it kind of confused me. So I tried the code above with both pyqrcode and qrcode, but both fail at the second line saying AttributeError: 'module' object has no attribute 'Decoder'. Furthermore, the website refers to Ubuntu 8.10 (which came out more than 6 years ago) and I can't find a public (git or other) repository of it to check the latest commit. So I moved on to the next library:

ZBar (website here) claims to be "an open source software suite for reading bar codes from various sources, such as image files." So I tried installing it on Mac OSX running sudo pip install zbar. This fails with error: command 'cc' failed with exit status 1. I tried to suggestions in the answers to this SO question, but I can't seem to solve it. So I decided to move on again:

QRTools, which according to this blogpost can decode images easily by using the following code:

from qrtools import QR
myCode = QR(filename=u"/home/psutton/Documents/Python/qrcodes/qrcode.png")
if myCode.decode():
  print myCode.data
  print myCode.data_type
  print myCode.data_to_string()

So I tried installing it using sudo pip install qrtools, which can't find anything. I also tried it with python-qrtools, qr-tools, python-qrtools and a couple more combinations, but unfortunately to no avail. I suppose it refers to this repo which says it is based on ZBar (see above). Although I want to run my code on Heroku (and thus prefer a pure Python solution) I successfully installed it on a Linux box (with sudo apt-get install python-qrtools) and tried running it:

from qrtools import QR
c = QR(filename='/home/kramer65/qrcode.jpg')
c.data  # prints u'NULL'
c.data_type  # prints u'text'
c.data_to_string()  # prints '\xef\xbb\xbfNULL' where I expect an int (being `1234567890`)

Although this seems to decode it, It doesn't seem to do it correctly. It furthermore needs ZBar and is thus not pure Python. So I decided to find yet another library.

PyXing (website here) is supposedly a Python port of the popular Java ZXing library, but the initial and only commit is 6 years old and the project has no readme or documentation whatsoever.

For the rest I found a couple qr-encoders (not decoders) and some API endpoints which can decode for you. Since I don't like this service to be dependent on other API endpoints I would want to keep the decoding local though.

So to conclude; would anybody know how I can decode QR-codes from images in (preferable pure) Python? All tips are welcome!

Answer 1

You can try the following steps and code using qrtools:

• Create a qrcode file, if not already existing

  • I used pyqrcode for doing this, which can be installed using pip install pyqrcode

  • And then use the code:

    >>> import pyqrcode
    >>> qr = pyqrcode.create("HORN O.K. PLEASE.")
    >>> qr.png("horn.png", scale=6)
    

• Decode an existing qrcode file using qrtools

  • Install qrtools using sudo apt-get install python-qrtools

  • Now use the following code within your python prompt

    >>> import qrtools
    >>> qr = qrtools.QR()
    >>> qr.decode("horn.png")
    >>> print qr.data
    u'HORN O.K. PLEASE.'
    

Here is the complete code in a single run:

In [2]: import pyqrcode
In [3]: qr = pyqrcode.create("HORN O.K. PLEASE.")
In [4]: qr.png("horn.png", scale=6)
In [5]: import qrtools
In [6]: qr = qrtools.QR()
In [7]: qr.decode("horn.png")
Out[7]: True
In [8]: print qr.data
HORN O.K. PLEASE.

Caveats

• You might need to install PyPNG using pip install pypng for using pyqrcode

• In case you have PIL installed, you might get IOError: decoder zip not available. In that case, try uninstalling and reinstalling PIL using:

  pip uninstall PIL
  pip install PIL
  

• If that doesn't work, try using Pillow instead

  pip uninstall PIL
  pip install pillow
  

Answer 2

For Windows using ZBar

Pre-requisites:

• Install ZBar by either:
  • Install Chocolatey and choco install zbar
  • Or use Windows Installer for ZBar
• pip install pyzbar

To decode:

from PIL import Image
from pyzbar import pyzbar

img = Image.open('My-Image.jpg')
output = pyzbar.decode(img)
print(output)

---

Alternatively, you can also try using ZBarLight by setting it up as mentioned here:
https://pypi.org/project/zbarlight/

Answer 3

There is a library called BoofCV which claims to better than ZBar and other libraries.
Here are the steps to use that (any OS).

Pre-requisites:

• Ensure JDK 14+ is installed and set in $PATH
• pip install pyboof

Class to decode:

import os
import numpy as np
import pyboof as pb

pb.init_memmap() #Optional

class QR_Extractor:
    # Src: github.com/lessthanoptimal/PyBoof/blob/master/examples/qrcode_detect.py
    def __init__(self):
        self.detector = pb.FactoryFiducial(np.uint8).qrcode()
    
    def extract(self, img_path):
        if not os.path.isfile(img_path):
            print('File not found:', img_path)
            return None
        image = pb.load_single_band(img_path, np.uint8)
        self.detector.detect(image)
        qr_codes = []
        for qr in self.detector.detections:
            qr_codes.append({
                'text': qr.message,
                'points': qr.bounds.convert_tuple()
            })
        return qr_codes

Usage:

qr_scanner = QR_Extractor()
output = qr_scanner.extract('Your-Image.jpg')
print(output)

Tested and works on Python 3.8 (Windows & Ubuntu)

Answer 4

The following code works fine with me:

brew install zbar
pip install pyqrcode
pip install pyzbar

For QR code image creation:

import pyqrcode
qr = pyqrcode.create("test1")
qr.png("test1.png", scale=6)

For QR code decoding:

from PIL import Image
from pyzbar.pyzbar import decode
data = decode(Image.open('test1.png'))
print(data)

that prints the result:

[Decoded(data=b'test1', type='QRCODE', rect=Rect(left=24, top=24, width=126, height=126), polygon=[Point(x=24, y=24), Point(x=24, y=150), Point(x=150, y=150), Point(x=150, y=24)])]

Answer 5

I'm answering only the part of the question about zbar installation.

I spent nearly half an hour a few hours to make it work on Windows + Python 2.7 64-bit, so here are additional notes to the accepted answer:

• Download https://github.com/NaturalHistoryMuseum/ZBarWin64/releases/download/v0.10/zbar-0.10-cp27-none-win_amd64.whl

• Install it with pip install zbar-0.10-cp27-none-win_amd64.whl

• If Python reports an ImportError: DLL load failed: The specified module could not be found. when doing import zbar, then you will just need to install the Visual C++ Redistributable Packages for VS 2013 (I spent a lot of time here, trying to recompile unsuccessfully...)

• Required too: libzbar64-0.dll must be in a folder which is in the PATH. In my case I copied it to "C:\Python27\libzbar64-0.dll" (which is in the PATH). If it still does not work, add this:

  import os
  os.environ['PATH'] += ';C:\\Python27' 
  import zbar
  

PS: Making it work with Python 3.x is even more difficult: Compile zbar for Python 3.x.

PS2: I just tested pyzbar with pip install pyzbar and it's MUCH easier, it works out-of-the-box (the only thing is you need to have VC Redist 2013 files installed). It is also recommended to use this library in this pyimagesearch.com article.

Answer 6

I found a new and effective way, just using cv2. The code below will decode a QR code.

import cv2
# Name of the QR Code Image file
filename = "attandence_Record_QR_code.png"
# read the QRCODE image
image = cv2.imread(filename)
# initialize the cv2 QRCode detector
detector = cv2.QRCodeDetector()
# detect and decode
data, vertices_array, binary_qrcode = detector.detectAndDecode(image)
# if there is a QR code
# print the data
if vertices_array is not None:
  print("QRCode data:")
  print(data)
else:
  print("There was some error") 

Answer 7

first intsall pip install pyzbar

then:

import cv2 as cv
from pyzbar.pyzbar import decode
     img=cv.imread('/path/img.jpg')
     objs=decode(img)
     for obj in objs:
         print('data: ', obj) 

Answer 8

PyBoof is a wrapper around BoofCV and has an easy to use QR Code reader. It also give you access to a ton of information about the QR, e.g. number of bit errors, precise location, raw message, ...

image = pb.load_single_band(data_path, np.uint8)

detector = pb.FactoryFiducial(np.uint8).qrcode()
detector.detect(image)

for qr in detector.detections:
    print("Message: " + qr.message)
    print("     at: " + str(qr.bounds))

Answer 9

In 2022 using Python 3 this is what worked for me.

Step 1: Install the zbar library

macOS:

brew install zbar

If you get an 'Unable to find zbar shared library' ImportError then do:

mkdir ~/lib
ln -s $(brew --prefix zbar)/lib/libzbar.dylib ~/lib/libzbar.dylib

Ubuntu/Debian Linux:

sudo apt-get update
sudo apt-get install zbar-tools

Windows:

According to the pyzbar project page: Install "Visual C++ Redistributable Packages for Visual Studio 2013. Install vcredist_x64.exe if using 64-bit Python, vcredist_x86.exe if using 32-bit Python."

Step 2: Install the PIP Python packages

pip3 install opencv-python
pip3 install pyzbar

Step 3: Create a Python script

Create a .py script and run the following:

import cv2 as cv
from pyzbar.pyzbar import decode

qrcode_img = 'path/to/qrcode/img.png'
img = cv.imread(qrcode_img)
decoded_data = decode(img)

# parse the decoded zbar data
url = decoded_data[0].data.decode()
print(f'URL: {url}')