Barcode Capture

Introduction

This code was made to extract any barcode using your web camera. As of now it utilizes pyzbar and pylibdmtx libraries. Which have their own set of recognizable barcodes. Mainly, we were targeting QR codes and data matrix codes. But it will recognize other popular formats supported by those libraries.

Under the hood, this code takes each frame from the camera like so

And detects barcode if it present. And finally provides output as a string like so

output: Wikipedia, the free encyclopedia

There is a probability that algorithm will fail during one of those steps. In such case the code will not produce any output. Also, there is a small possibility that the algorithm will provide incorrect output. So keep that in mind while capturing barcodes!

Requirements

This code has been tested on Windows machine. Most likely it will work on alternative OS but we have not tested it on any other OS.

Otherwise, you will need:

• Python 3
• OpenCV 2
• pyzbar
• pylibdmtx
• Web Camera

Easy installation

Use pip to install this library, fast and easy!!!

pip install barcap

Manual installation

Install libraries

On Linux, you have to install some extra packages:

sudo apt-get install libzbar0
sudo apt-get install libdmtx0a

On MAC, that be a little different:

brew install zbar
brew install libdmtx

On Windows, everything should happen automatically. If you have any dll problems, please install Visual C++ Redistributable Packages for Visual Studio 2013 and see if that resolves those issues.

Install python packages

You can install requirements using

pip install -r requirements.txt

Further instructions

• https://pypi.org/project/pyzbar/
• https://pypi.org/project/pylibdmtx/

Most Basic Barcode Capture

This is the simplest way to get you started. Also this is the best way to make sure that your setup works properly. Simply run in a shell:

python -m barcap

Window with your default camera is going to pop up and capture algorithm will be running in a while loop. If everything is successful you will see output periodically printed out in the shell.

Manual Capture

Once again start while capture loop by running:

$ python -m barcap

First of all, make sure your capture window is selected. Once, you have a good image in the window with barcode occupying most of the screen press 's' key to save this frame. Open barcode.jpg and confirm that the frame has been written. Kill the loop by pressing 'q' or Esc.

If it does not work...

• Do you have camera connected?
• Is computer even on?
• If nothing helps fix your hardware in the software!

All jokes aside most likely the libraries do not support this barcode or something wrong with a setup.

Final Thoughts

At this point you probably understand how my routine works so you can try modifying it if things are not working out.

Otherwise, Congratulations!!! You got it working!!!

Importing Barcode Capture into your own python script

If you look closely, you will see that capture loop is inside CaptureProcess class based on Process class from multiprocessing module.

The idea is to run this process and capture frames from a web camera. Next, those frames will be processed to get desired output.

You could use this class to build your own capturing algorithm. For examples, look closely at barcode.py, ocr.py and ocr_plus.py.

And finally, the user of the capture process would start a process and get results via shared memory.

Here is how you can take advantage of those in your own script.

To create a capture instance do following:

from barcap.barcode import BarcodeCapture     
capture = BarcodeCapture(camera=0)

To start capturing process:

capture.start()

To stop capturing process:

capture.stop()

or

capture.terminate()

To check if capturing is still happening:

capture.is_alive()

To read decoded output:

output = capture.ouput
print(f'output: {output}')

To fetch epoch time of the last decoding:

import time

epoch = capture.last_epoch
time_stamp = time.strftime('%Y-%m-%d %H:%M:%S', time.localtime(epoch))
print(f'last capture: {time_stamp}')

You can also check if new output is available

if capture.new:
    print(f'output: {capture.output}')

Every time you read output it resets the new flag.

Look closely at main.py to get a complete example.

Finally, device_list.py facilitates in selecting right camera. In case you might have multiple cameras connected to your computer.

Device List

There is a little device utility that can help you discover connected cameras to the computer. As of right now the sub package works only on Windows. Also, we decided that this is optional addition so it won't be compiled by default.

In order to compile navigate to setup.py script and uncomment lines for the WindowsDevice extension. Reinstall barcap using setup.py script.

Additional Info and Questions

Shoot me email at kirill at kbelyayev.com if you have any questions, suggestions, improvements, additions and etc. I would love to help you with this script if you hire me as a contractor. I might help you free of charge if you contribute to this distribution or ask politely. Beer donations are welcome too!

Good luck! Happy coding and hacking!