PDF Form Extraction Tool

Overview

This Python script extracts information from PDF forms using OCR (Optical Character Recognition) and saves the extracted data into an Excel file. It is particularly designed for processing forms with checkboxes and textual fields. The script can handle variations in form structure and allows for easy customization to accommodate other PDF form types.

---

Features

• Extracts textual data and checkbox states from PDF forms.
• Uses Tesseract OCR for text recognition.
• Detects and processes checkboxes to determine their states (checked/unchecked).
• Saves extracted information in a structured Excel format.
• Configurable field mappings to adapt to different PDF forms.

---

Prerequisites

Python Libraries

Install the required libraries by running:

pip install numpy opencv-python pytesseract PyMuPDF==1.20.2 pillow pandas openpyxl

Tesseract OCR

Download and install Tesseract OCR from Tesseract OCR GitHub.

Update the pytesseract.pytesseract.tesseract_cmd variable in the script to the path of the Tesseract executable on your system.

Example:

pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"

---

Usage

1. Place the PDF File Ensure the input PDF form is located in the specified path (e.g., input folder/Completed Dispute Form 3.pdf).

2. Run the Script Execute the script with:

   python script_name.py

   Replace script_name.py with the actual filename of the script.

3. Output The extracted data will be saved in the specified Excel file path (e.g., output folder/extracted_data_form3.xlsx).

---

Customization for Different PDF Forms

Modify Field Mappings

To adapt the script for different forms:

• Update the FIELDS dictionary to include all the fields you want to extract.
• Modify the field_mappings dictionary to map field names in the form to keywords/phrases that identify them in the text.

Example:

field_mappings = {
    "New Field Name": ["Keyword or phrase in the form"],
    ...
}

Adjust Checkbox Detection

• Place your .pdf files in cb_test.py and run it to verify whether the checkboxes in your files are getting recognized.
• If your check-boxes arent getting recognized properly, try adjusting values min_size, max_size, min_fill_ratio.
• Update the logic in the handle_page_1_checkboxes and handle_page_2_checkboxes functions to account for new checkbox labels or states.

PDF Page Handling

If the form has additional pages or layouts:

• Extend the update_fields_from_checkbox_results function to handle new page numbers or checkbox patterns.
• If you want to utilize individual components of the process, u can access in the Project folder and integrate it into your own process.

---

Script Flow

1. PDF Parsing
   • Converts PDF pages to images using PyMuPDF.
2. Text and Checkbox Extraction
   • Extracts text using Tesseract OCR.
   • Detects and processes checkboxes based on contours and fills.
3. Field Matching
   • Matches extracted data against predefined fields and mappings.
4. Output to Excel
   • Saves results to an Excel file using Pandas.

---

Example Output

Extracted data is saved in an Excel file with two columns:

• Field Name: Name of the extracted field.
• Extracted Value: Corresponding value from the form.

---

Contribution

Feel free to contribute by:

• Reporting issues.
• Adding support for new form layouts.
• Improving OCR accuracy or field mapping logic.

---

License

This project is open-source.