🐍 RPA Challenge OCR Invoice Extraction - Python

This project automates the process of extracting invoice details from the RPA Challenge OCR website: https://rpachallengeocr.azurewebsites.net/. It uses Playwright for web automation to interact with the website, and Tesseract OCR for optical character recognition (OCR) to extract text from invoice images. The extracted data is then structured into a CSV file and uploaded back to the website for processing.

Key Features:

• Automated Web Navigation: Uses Playwright to navigate and interact with the web pages.
• Invoice Image Extraction: Downloads invoice images from the website for further processing.
• OCR Processing: Tesseract OCR extracts text from the downloaded invoice images.
• Data Parsing: Extracts relevant invoice details using regex.
• CSV Generation: Saves extracted data in a structured format.
• Automated Submission: Uploads the CSV file back to the site.

---

Process Overview

1. Launch Browser:

• The program begins by launching a Chromium browser instance using Playwright.
• The browser navigates to the RPA Challenge OCR website, where it can interact with various elements to extract data.

2. Invoice Data Extraction:

• The program searches the website for invoices that are due for processing.
• For each invoice, the program clicks on the corresponding link to open the invoice details in a popup window.

3. Image Download:

• The invoice image URL is extracted from the webpage, and the image is downloaded using the requests library.
• The image is saved locally in the temp/invoices folder to be processed later. This ensures that images are organized and easily accessible.

4. OCR Processing:

• Tesseract OCR is used to extract text from the downloaded invoice image.
• The extracted text is then processed using regular expressions (regex) to identify the necessary details, such as:
  • Invoice Number: The unique identifier for the invoice.
  • Invoice Date: The date the invoice was issued.
  • Company Name: The name of the company issuing the invoice.
  • Total Due: The total amount due on the invoice.

5. Data Structuring:

• The extracted data is stored in a structured format using pandas DataFrame.
• The DataFrame is then converted into a CSV file, which contains the following columns:
  • Invoice ID: Unique ID of the invoice.
  • Due Date: The due date of the invoice.
  • Invoice Number: Extracted invoice number.
  • Invoice Date: Extracted invoice date.
  • Company Name: Extracted company name.
  • Total Due: Extracted total due amount.
• The generated CSV file is saved in the temp folder, specifically in the temp/invoices.csv file.

6. Upload Data:

• The generated CSV file is automatically uploaded back to the website for processing. This is done by interacting with the appropriate input field on the site using Playwright.

---

Tech Stack and Dependencies

This project utilizes Python and the following key libraries:

Package	Purpose
playwright	Automates browser interactions for web scraping and automation.
pytesseract	Performs OCR to extract text from invoice images.
pandas	Used for data processing, structuring, and saving the extracted data in CSV format.
Pillow	Provides image processing utilities (used for opening and handling images).
requests	Downloads invoice images from URLs.

Dependencies Installation:

To install the required dependencies, make sure you have Python installed, then use the following command to install them:

pip install -r requirements.txt

---

License

This project is licensed under the MIT License. Please see the LICENSE file for more details.