Video to PDF Converter

Overview

This project focuses on extracting pages from a video where a book's pages are flipped one by one. The captured frames with visible pages are then used to create a PDF containing all the book pages.

Project Workflow

1. Video Input: Provide a video file (*.mp4) where a book's pages are being flipped. The video can have some hand motion and camera movements.

2. Page Capture:

   • Detect significant changes between consecutive frames to identify page flips.
   • Capture frames when a page flip is detected and when the page is stable for better visibility.

3. Page Extraction:

   • Use computer vision techniques to extract the page from the captured frames.
   • Remove background noise and distortion to isolate the page.

4. PDF Generation:

   • Save the extracted pages as images.
   • Create a PDF containing all the extracted pages.

Usage

1. Clone this repository to your local machine:

   git clone https://github.com/yourusername/book-page-extraction.git
   

2. Install the required libraries:

   pip install opencv-python-headless reportlab numpy Pillow

3. Place your video file (*.mp4) in the project directory.

4. Modify the video_path variable in main.py to point to your video file.

5. Run the script: pyton video_to_pdf.py

6. Once the script finishes, you will find the extracted pages in the project directory as JPEG images (saved_frame_*.jpg) and a PDF file (content_pdf_1.pdf) containing all the pages

Dependencies

• OpenCV: For computer vision tasks such as frame extraction and page detection.
• ReportLab: For generating PDF files.
• NumPy: For numerical operations.
• Pillow: For image processing.

Notes

• Adjust the threshold values and parameters in the code according to the specific characteristics of your video.
• Ensure proper lighting and stable camera setup for better results.
• For detailed explanations and optimizations of the code, refer to the video_to_pdf.py file.