How to Scan Barcodes in Jupyter Notebook

Dec 19, 2021

Jupyter notebook is a web-based interactive computing environment that enables you to create and share documents that contain live code, equations, visualizations and narrative text. If you have installed Anaconda, Jupyter notebook is installed by default. This article demonstrates how to write Python code to scan barcodes in Jupyter notebook.

Pre-requisites

Apply for a trial license of Dynamsoft Barcode Reader SDK.

Creating a Jupyter Notebook for Python Barcode Scanning

Run Jupyter notebook:
Open http://localhost:8888/ in your browser to create a barcode_scanning.ipynb file.
Insert a new cell to install Dynamsoft Barcode Reader, OpenCV, and Matplotlib:
```
 %pip install dynamsoft-barcode-reader-bundle opencv-python matplotlib
```

Initialize the license of barcode SDK and barcode reader object:

 from dynamsoft_capture_vision_bundle import *
 license_key = "LICENSE-KEY"
 error_code, error_message = LicenseManager.init_license(license_key)
 cvr_instance = CaptureVisionRouter()

Create an image upload button in Jupyter notebook:

 from ipywidgets import FileUpload
 def on_upload_change(change):
     if change.name != "value":
         return
     if not change.new:
         return
    
     up = change.owner
     files_obj = change.new
    
     records = list(_flatten_records(files_obj))
    
     for rec in records:
         try:
             content = _get_bytes(rec)  
         except Exception as e:
             print("Skip unsupported record:", e)
             continue
    
         buf = np.frombuffer(content, dtype=np.uint8)
         img = cv.imdecode(buf, cv.IMREAD_COLOR)
         if img is None:
             print("Failed to decode image.")
             continue
    
         new_img = img.copy()
         try:
             decode(new_img)
         except Exception as e:
             print("Decode error:", e)
    
         img_rgb = cv.cvtColor(img, cv.COLOR_BGR2RGB)
         new_rgb = cv.cvtColor(new_img, cv.COLOR_BGR2RGB)
         show_image(img_rgb, new_rgb)
    
     try:
         up.unobserve(on_upload_change, names="value")
     except Exception:
         pass
    
     reset_done = False
     for val in ((), [], None):
         try:
             up.value = val
             reset_done = True
             break
         except Exception:
             continue
     if not reset_done:
         try:
             up.value.clear()
             reset_done = True
         except Exception:
             pass
    
     try:
         up.observe(on_upload_change, names="value")
     except Exception:
         pass
    
 uploader = FileUpload(accept='image/*', multiple=True) 
 uploader.observe(on_upload_change, names="value")
 uploader

Convert the image data to numpy array and then decode the image with OpenCV API:

 import cv2 as cv
 import numpy as np

 for rec in records:
     try:
         content = _get_bytes(rec)  
     except Exception as e:
         print("Skip unsupported record:", e)
         continue

     buf = np.frombuffer(content, dtype=np.uint8)
     img = cv.imdecode(buf, cv.IMREAD_COLOR)

Scan barcodes with Dynamsoft Barcode Reader API:

 def decode(frame):
     before = time.time()
     result = cvr_instance.capture(frame, EnumPresetTemplate.PT_READ_BARCODES.value)
     after = time.time()
    
     COLOR_RED = (0, 0, 255)
     thickness = 2
     text_x, text_y = 10, 20
    
     if result.get_error_code() != EnumErrorCode.EC_OK:
         print("Error:", result.get_error_code(), result.get_error_string())
         cv.putText(frame, f'{after - before:.2f} s, barcode found: 0', (text_x, text_y),
                    cv.FONT_HERSHEY_SIMPLEX, 0.5, COLOR_RED)
         return
    
     items = result.get_items()
     found = len(items)
     for item in items:
         format_type = item.get_format()
         text = item.get_text()
         print(f'Format: {format_type}, Text: {text}')
         loc = item.get_location()
         pts = np.array([[p.x, p.y] for p in loc.points], dtype=np.int32)
         cv.drawContours(image=frame, contours=[pts], contourIdx=-1,
                         color=COLOR_RED, thickness=thickness, lineType=cv.LINE_AA)
         cv.putText(frame, f'{format_type}', (pts[0][0], pts[0][1] - 10),
                    cv.FONT_HERSHEY_SIMPLEX, 0.5, COLOR_RED)
         cv.putText(frame, f'{text}', (pts[0][0], pts[0][1] - 30),
                    cv.FONT_HERSHEY_SIMPLEX, 0.5, COLOR_RED)
     cv.putText(frame, f'{after - before:.2f} s, barcode found: {found}', (text_x, text_y),
                cv.FONT_HERSHEY_SIMPLEX, 0.5, COLOR_RED)
        
 new_img = img.copy()
 decode(new_img)

Display the barcode recognition results with Matplotlib:

 import matplotlib.pyplot as plt

 def show_image(img1, img2):
     fig = plt.figure(figsize=(18, 8))
     ax1 = fig.add_subplot(1, 2, 1)
     ax1.set_title('Input image', fontsize=14)
     ax1.axis('off')
     ax2 = fig.add_subplot(1, 2, 2)
     ax2.set_title('Barcode Recognition', fontsize=14)
     ax2.axis('off')
     ax1.imshow(img1)
     ax2.imshow(img2)
     plt.show()

Scan barcodes in Jupyter Notebook

Source Code

https://github.com/yushulx/python-barcode-qrcode-sdk/tree/main/examples/official/jupyter_notebook