Transform Your Images into Pencil Sketches with Python 🚀

Have you ever wondered how to create a pencil sketch effect from an image programmatically? With Python and some powerful libraries, you can achieve this in just a few lines of code! In this article, I'll walk you through a simple script that transforms any image into a pencil sketch using grayscale conversion, Gaussian blur, and a dodge blending effect.

---

📜 The Concept Behind Pencil Sketching

The pencil sketch effect relies on simulating how light and shadow interact in hand-drawn sketches. Here's the breakdown:

1. Convert the image to grayscale: Remove all colors to simplify the image to shades of gray.
2. Invert the grayscale image: Create a negative effect where light becomes dark and vice versa.
3. Apply a Gaussian blur: Blur the inverted image to soften details.
4. Dodge blending: Blend the blurred image with the original grayscale to simulate pencil strokes.

---

🚀 Python Implementation

Here's the Python code that does it all:

import numpy as np
import imageio.v2 as imageio  # Explicit use of imageio.v2
import cv2
import scipy.ndimage


def convert_to_grayscale(rgb_image):
    """
    Converts an RGB image to grayscale.
    """
    return np.dot(rgb_image[..., :3], [0.299, 0.587, 0.114])


def apply_dodge(front_layer, back_layer):
    """
    Applies the 'dodge' effect to create a pencil sketch effect.
    """
    result = front_layer * 255 / (255 - back_layer)
    result[result > 255] = 255
    result[back_layer == 255] = 255
    return result.astype('uint8')


def pencil_sketch(image_path, output_path, blur_sigma=10):
    """
    Generates a pencil sketch effect from an image.

    Args:
        image_path (str): Path to the input image.
        output_path (str): Path to the output image.
        blur_sigma (float): Sigma for the Gaussian blur (defines the blur level).
    """
    # Load the image
    original_image = imageio.imread(image_path)

    # Convert to grayscale
    grayscale_image = convert_to_grayscale(original_image)

    # Invert the colors
    inverted_image = 255 - grayscale_image

    # Apply Gaussian blur
    blurred_image = scipy.ndimage.gaussian_filter(inverted_image, sigma=blur_sigma)

    # Apply the dodge effect
    final_image = apply_dodge(blurred_image, grayscale_image)

    # Save the resulting image
    cv2.imwrite(output_path, final_image)
    print(f"The image has been saved to '{output_path}'.")


# Example usage
if __name__ == "__main__":
    input_image = "test.jpeg"
    output_image = "test_coloring.png"
    pencil_sketch(input_image, output_image)

---

🔍 Understanding the Functions

1. Grayscale Conversion

The function convert_to_grayscale uses a weighted sum to convert an RGB image into grayscale. This replicates human perception, where green contributes the most, followed by red, and then blue.

np.dot(rgb_image[..., :3], [0.299, 0.587, 0.114])

2. Dodge Effect

The apply_dodge function lightens the grayscale image, mimicking pencil strokes. It ensures pixel values stay within valid bounds (0-255).

def apply_dodge(front_layer, back_layer):
    result = front_layer * 255 / (255 - back_layer)
    result[result > 255] = 255
    result[back_layer == 255] = 255
    return result.astype('uint8')

3. Gaussian Blur

The blur effect softens the details of the inverted image, helping create smoother pencil strokes. It is achieved using the scipy.ndimage.gaussian_filter function, where the sigma parameter controls the intensity of the blur.

blurred_image = scipy.ndimage.gaussian_filter(inverted_image, sigma=blur_sigma)

---

🖼️ Example Usage

Save the script as pencil_sketch.py, then run it with the following setup:

python main.py

Input:

Place an image (e.g., test.jpeg) in the same directory.

Output:

The output will be saved as test_coloring.png, a beautiful pencil sketch version of the input image.

---

🛠️ Tips for Customization

• Adjust Blur Intensity: The blur_sigma parameter controls the level of detail in the sketch. A lower value retains more details, while a higher value creates a smoother sketch.

pencil_sketch("input.jpg", "output.png", blur_sigma=15)

• Error Handling: Add checks to handle invalid paths or unsupported formats.

if not os.path.exists(image_path):
    raise FileNotFoundError("The input image file does not exist.")

• Performance Boost: Use OpenCV's cv2.cvtColor for faster grayscale conversion if performance is critical.

---

🌟 Resources

• NumPy Documentation
• Imageio Documentation
• OpenCV Documentation
• SciPy Documentation

---

🔮 Conclusion

This script showcases how Python's versatile libraries can simplify complex image processing tasks. Whether you're a beginner or a seasoned developer, transforming images into pencil sketches is a fun and rewarding project. So, why not try it out with your favorite pictures ?