Bjorn score

Calculate image similarity score

Tools used:

Language: Python
Dev Modules Used: 
- opencv-python
- scikit-image
- click
Test Modules Used:
- tox
- pytest
- pylint
- pytest-cov
Build and Deployment: Travis-ci
App Repository: PyPi 

Installation:

pip3 install imagescore

How to run:

imagescore --infile 'input.csv' --outfile 'output.csv'

The options are:

    --infile  | -i [str]    -- [Input file path]
    --outfile | -o [str]    -- [Output file path]
    --height  | -h [int]    -- [Optional: height to be resized, default = 4096]
    --width   | -w [int]    -- [Optional: width to be resized, default = 4096]

Expected Input: csv file with images and its absolute path
Expected Output: csv file with images and its absolute path, image score and elapsed time in secs.

Developer mode:

The application is written in python3. Hence create a virtualenv with python3 and install the dependencies from requirements.txt.

virtualenv --python=python3 venv
pip install -e .

The unit test cases are located in tests directory. Install the dependencies requrements-dev.txt.

pip install -r requirements-dev.txt
tox

Deployment:

The python package is build with the travis.yml script when there are changes on the master branch. To push a new version: Make the changes in the application, push it to any feature branch and merge with master.
To deploy a new version tag it. e.g

git tag -a v1.1.0
git push origin v1.1.0

When a tag is pushed, travis starts building and uploads the application to pypi
Pypi Release: https://pypi.org/project/imagescore/
Github Release: https://github.com/sandjaie/image_score/releases
Travis Builds: https://travis-ci.com/sandjaie/image_score/

Manual Deployment:
We are using twine to upload the artifact to Pypi registry. Pypi registry needs an account to be created. Enter the credentials in the cli when prompted.

pip3 install twine
python3 setup.py sdist bdist_wheel
twine upload dist/*