sathiyajith / electronic-invoicing

Document extraction for processing scanned invoices.

Electronic-Invoicing

Scanned invoices are extracted using image processing to reduce the non-reliability and man-power in calculating payments and invoice bills. Additionally, the whole payment process is automated with less turn around time and more flexibility in the invoice templates.

Deliverables (as promised) 😉

1. Accurate Prediction
2. Reducing cost of time and space
3. Automating changes and cycles
4. Reliability of data is ensured after every updation
5. Ease of access
6. Standard Invoice Format ensured
7. Scalable to any structured and unstructured documents

A detailed explanation of the CODE and the EXECUTION is available at E-Invoicing

Instructions (Let's get going)

RECOMMENDED: Running in GOOGLE COLAB provides faster results and interactivity

Directory Structure (Initial SETUP for COLAB & LOCALHOST)

/flask.py
/pdfconvert.py
/digital_processing.py
/table_extraction.py
/text_extraction.py
/templates
  /index.html
/static
  /index.css
/output
  /dataset1/
    invoice.pdf

Important Points

1. static & templates are required for FLASK operation.
2. Place the TEST INVOICE inside output/dataset1.
3. .py files in the root folder are responsible for the extraction processes.
4. At any time during the entire running of the program, intermediate O/P's are present at output/dataset1.
5. Every function present in the CODE has a clear DOCSTRING attached to it which can be called using help(function-name) (or) function-name.__doc__

Running in GOOGLE COLAB

You can either directly mount this Colab drive link in your colab, change path according to the project structure from your GDrive in the code and run flask.py or follow the below mentioned steps.

1. Ensure the above directory structure is maintained by moving all the 5 .ipynb modules and 2 folders static and templates.
2. Connect to the runtime environment and mount the GDrive.
3. RUN ALL Cells in flask.ipynb to fire up the WEB SERVER 🎉
4. Upload the invoice
5. Initially the CONVERT button is blocked and after pre-processing it is enabled.
6. The final O/P's are available at output/dataset1 including pre-processing steps. (In case the process is slow please refer to this directory)
7. Finally .zip containing all the required information is available at output/dt1.zip

Running in LOCALHOST

1. Clone the repo and extract the localhost code to a seperate folder.
2. pip install -r requirements.txt located here
3. Run python index.py inside the directory and you're good to go 🎉
4. Upload the invoice
5. Initially the CONVERT button is blocked and after pre-processing it is enabled.
6. The final O/P's are available at output/dataset1 including pre-processing steps. (In case the process is slow please refer to this directory)
7. Finally .zip containing all the required information is available at output/dt1.zip

If there are any issues, I request you to mailto

electronic-invoicing's Issues

Documentation

Documentation and other minor fixes needed