Image Conversion to Grayscale and Binarization

This repository contains a simple project demonstrating how to:

1. Load a JPEG image (using the Pillow library).
2. Convert the image to grayscale, pixel by pixel, without using any built-in conversion methods.
3. Binarize the image using a manual threshold.
4. Display the three resulting images (original, grayscale, and binarized) side by side, using Matplotlib.

---

Project Structure

The notebook (or script) is generally composed of the following parts:

1. Imports and Environment Check

   • Installs and/or imports the necessary libraries: Pillow (for image I/O), matplotlib (for displaying images), and os (for path checks).
   • Example:

     # If Pillow is not installed:
     # !pip install Pillow
     
     from PIL import Image
     import matplotlib.pyplot as plt
     import os

2. Function: convert_to_grayscale_and_binary(image_path, threshold=128)

   • This function reads an image file from the given path (image_path).
   • It then manually converts each pixel from RGB to a single grayscale value using the approximate NTSC formula:

     gray_value = 0.299 * R + 0.587 * G + 0.114 * B
     

   • Subsequently, it applies a threshold (default = 128) to convert any pixel above or equal to this threshold into white (255), and anything below into black (0).
   • Returns two Pillow Image objects: one grayscale and one binarized.

3. Main Execution Steps

   • Defines the path to the original image (e.g., /content/lena.jpeg).
   • Checks if the file exists.
   • Calls the convert_to_grayscale_and_binary() function to obtain:
     • A grayscale image.
     • A binarized image.
   • Saves those images to disk, e.g.:

     img_gray.save("gray_output.png")
     img_bin.save("binary_output.png")
     

   • Uses Matplotlib to display the three images side by side: original, grayscale, and binarized.

   Example usage:

   image_path = "/content/lena.jpeg"
   img_gray, img_bin = convert_to_grayscale_and_binary(image_path, threshold=128)

---

Requirements

• Python 3.x
• Pillow (PIL): for image manipulation (loading and saving).
• Matplotlib: for displaying images.
• OS: only if you need to perform file/path checks (e.g., verifying if the file exists).
• Jupyter Notebook or Google Colab (optional), if you want to run the code interactively.

Installation instructions:

pip install Pillow matplotlib

---

Usage

1. Clone or download this repository (or copy the relevant notebook/script).
2. Install the required packages:

   pip install Pillow matplotlib

3. Place your desired image in the same directory or specify the correct path in the code. For example:

   image_path = "lena.jpeg"

4. Run the notebook or Python script. You should see:
   • A grayscale version of the original image.
   • A binarized version of the original image (black and white).
   • The three versions displayed side by side in your output or notebook.

---

Code Explanation

convert_to_grayscale_and_binary(image_path, threshold=128)

• Input:
  • image_path (string): path to the input image.
  • threshold (int, optional): cutoff value for binarization (default = 128).
• Process:
  1. Opens the image with Image.open(image_path).convert('RGB').
  2. Creates two new images: one for grayscale (mode 'L') and one for binary (also mode 'L').
  3. Iterates through every pixel in the original image, calculates gray_value, and then applies the threshold.
  4. Sets each pixel in the respective output images.
• Returns:
  • grayscale_img (PIL.Image): image in 8-bit grayscale.
  • binarized_img (PIL.Image): image in black and white (0 or 255).

Displaying Images

• Matplotlib usage:

  import matplotlib.pyplot as plt
  
  plt.figure(figsize=(12, 4))
  
  # Original
  plt.subplot(1, 3, 1)
  plt.imshow(img_original)
  plt.title("Original")
  plt.axis('off')
  
  # Grayscale
  plt.subplot(1, 3, 2)
  plt.imshow(img_gray, cmap='gray')
  plt.title("Grayscale")
  plt.axis('off')
  
  # Binarized
  plt.subplot(1, 3, 3)
  plt.imshow(img_bin, cmap='gray')
  plt.title("Binarized")
  plt.axis('off')
  
  plt.tight_layout()
  plt.show()

This approach allows a clear side-by-side visualization of the transformations.

---

Contributing

Feel free to open issues, suggest improvements, or add new examples (e.g., dynamic threshold selection) via pull requests.

---

License

This project is provided under the MIT License. You are free to use, modify, and distribute this code as permitted by the license.