docker-tesserocr

A docker container based on hmmbug/tesseract with Tesseract 4.00 (alpha), Python 2.7 and the tesseroct python bindings.

This container has been produced for the Digital Bangkok Recorder Project.

Usage

Run a python script

docker run --name tesserocr -v appdir:/work -w work hmmbug/tesserocr myapp.py arg1 ... argn

where:

'appdir' is the local directory containing the app code which will be mounted in the container
'myapp.py' is the script to run
arg1 ... argn are the arguments to provide to the script

Run an interactive shell

docker run --rm -ti hmmbug/tesserocr /bin/sh

Other options

docker run hmmbug/tesserocr [-h|-t]

-h Print the usage text
-t Run the test script

Shortcut

A shell alias can be a convenient way to run the container:

alias tesserocr='docker run --rm -v \`pwd\`:/work -w /work hmmbug/tesserocr'

Then just run:

$ tesserocr -t
hmmbug/tesserocr test script

Software versions:
     python: 2.7.13
  tesseract: 4.00.00alpha
  leptonica: leptonica-1.74
  tesserocr: 2.2.0-beta

   Image: hello_world.png
Expected: Hello world.
  Actual: Hello world.
  Result: OK (0.266s)

   Image: quick_fox.png
Expected: The quick brown fox jumps over the lazy dog
  Actual: The quick brown fox jumps over the lazy dog
  Result: OK (0.756s)

$

Test script

The supplied test script loads a list of image files and performs OCR on them and then checks the actual output from the expected output.

Current tests:

hello_world.png: Hello world.
quick_fox.png: The quick brown fox jumps over the lazy dog