v0.13.0

Features

• Improve logging for Keyword PDF Should Contain Strings
• Add Keyword PDF Should Not Contain Strings

v0.12.1: Merge pull request #74 from manykarim/fix_file_lock_on_error

Fix

• Fix PDF Document Locks an assertion error also for Pdf Tests

v0.12.0: Merge pull request #73 from manykarim/fix_file_lock_on_error

Fix

• PDF files are no longer locked after failed comparisons as described here #17

v0.11.0

Fix

• Prevent upscaling image for OCR if resolution would be too high for tesseract (>32767 pixel width or height)

v0.10.1

Get Barcodes From Document

• will return a dictionary like {'x':10, 'y':20, 'width':100, 'height':30} when using return_type=coordinates
• will return a list [values, coordinates] when using return_type=all

v0.10.0

New Features

More barcode types supported thanks to pyzbar.

EAN-13/UPC-A, UPC-E, EAN-8, Code 128, Code 39, Interleaved 2 of 5 and QR Code

On Linux additional package needs to be installed
apt-get install libzbar0

v0.9.1

Defect Fixes
Fixed a defect caused by conversion to int in CompareImage.py
Now the string value (from tesseract) is first interpreted as float and afterwards converted to int

Thank you for reporting it @wanghui-00

v0.9.0

New Features

• Add new movement_detection options classic, template and orb
  • Relevant when using move_tolerance option in Compare Images.
  • orb support is still experimental, see OpenCV Documentation for more details
  • classic will use the current default (which is templateMatching)
  • template has a few improvements like some additional preProcessing done on both images before
    • for now, I recommend to use template
• Improved text detection with ocr_engine=east
  • If you face problems detecting texts in your images, try out ocr_engine=east. It has some improvements.

Fixes

• #62
• #48
  • added an explicit conversion to float and to int in CompareImage.py, I hope that fixes it
  • int(float(d['conf'][j])) > self.PYTESSERACT_CONFIDENCE

v0.8.1

Fix Visual Comparison from URLs

A fix for #59
It is now possible to perform Visual Comparisons where both Reference and Candidate Image are provided as URLs

v0.8.0

Improve OCR

• Provide option ocr_engine in Get Text From Document Keyword
• Create black/white image when using east text detection plus tesseract
• Use PYTESSERACT_CONFIDENCE = 20 constant to filter out false text results