How can a specific tint be added to grayscale images in scikit-learn in Python?

Adding tints to grayscale images involves manipulating the RGB channel values to create color effects. In scikit-image (part of the scikit-learn ecosystem), we convert grayscale images to RGB format and apply color multipliers to achieve different tints.

Required Libraries

First, let's import the necessary modules ?

import matplotlib.pyplot as plt
from skimage import data, color
from skimage import io
import numpy as np

Loading and Converting Image

We'll use a sample image from scikit-image's dataset and convert it to grayscale ?

# Load sample image (you can replace with your own image path)
orig_img = data.chelsea()  # Sample cat image from scikit-image

# Convert to grayscale
grayscale_img = color.rgb2gray(orig_img)

# Convert grayscale back to RGB (3-channel) for tinting
rgb_img = color.gray2rgb(grayscale_img)

print(f"Original shape: {orig_img.shape}")
print(f"Grayscale shape: {grayscale_img.shape}")
print(f"RGB converted shape: {rgb_img.shape}")

Original shape: (300, 451, 3)
Grayscale shape: (300, 451)
RGB converted shape: (300, 451, 3)

Applying Color Tints

We define color multipliers for different tints and apply them to create colored versions ?

# Define color multipliers for different tints
red_tint = [1.2, 0.3, 0.3]      # Red tint
blue_tint = [0.3, 0.3, 1.2]     # Blue tint
yellow_tint = [1.0, 1.0, 0.2]   # Yellow tint
green_tint = [0.3, 1.2, 0.3]    # Green tint

# Apply tints by multiplying with color multipliers
red_tinted = np.clip(rgb_img * red_tint, 0, 1)
blue_tinted = np.clip(rgb_img * blue_tint, 0, 1)
yellow_tinted = np.clip(rgb_img * yellow_tint, 0, 1)
green_tinted = np.clip(rgb_img * green_tint, 0, 1)

Displaying Results

Let's visualize the original grayscale image alongside the tinted versions ?

# Create subplot to display images
fig, axes = plt.subplots(2, 3, figsize=(15, 10))

# Display original and tinted images
axes[0, 0].imshow(grayscale_img, cmap='gray')
axes[0, 0].set_title('Original Grayscale')
axes[0, 0].axis('off')

axes[0, 1].imshow(red_tinted)
axes[0, 1].set_title('Red Tint')
axes[0, 1].axis('off')

axes[0, 2].imshow(blue_tinted)
axes[0, 2].set_title('Blue Tint')
axes[0, 2].axis('off')

axes[1, 0].imshow(yellow_tinted)
axes[1, 0].set_title('Yellow Tint')
axes[1, 0].axis('off')

axes[1, 1].imshow(green_tinted)
axes[1, 1].set_title('Green Tint')
axes[1, 1].axis('off')

axes[1, 2].axis('off')  # Hide empty subplot

plt.tight_layout()
plt.show()

Custom Tint Function

Here's a reusable function to apply any custom tint to grayscale images ?

def apply_tint(grayscale_img, tint_color):
    """
    Apply a color tint to a grayscale image.
    
    Parameters:
    grayscale_img: 2D numpy array (grayscale image)
    tint_color: list of 3 values [R, G, B] for tint multipliers
    
    Returns:
    Tinted RGB image
    """
    # Convert grayscale to RGB
    rgb_img = color.gray2rgb(grayscale_img)
    
    # Apply tint and clip values to valid range
    tinted_img = np.clip(rgb_img * tint_color, 0, 1)
    
    return tinted_img

# Example usage
sepia_tint = [1.0, 0.8, 0.6]  # Sepia effect
purple_tint = [0.8, 0.4, 1.0]  # Purple tint

sepia_img = apply_tint(grayscale_img, sepia_tint)
purple_img = apply_tint(grayscale_img, purple_tint)

# Display results
fig, (ax1, ax2, ax3) = plt.subplots(1, 3, figsize=(12, 4))

ax1.imshow(grayscale_img, cmap='gray')
ax1.set_title('Original Grayscale')
ax1.axis('off')

ax2.imshow(sepia_img)
ax2.set_title('Sepia Tint')
ax2.axis('off')

ax3.imshow(purple_img)
ax3.set_title('Purple Tint')
ax3.axis('off')

plt.tight_layout()
plt.show()

Key Points

• Color Multipliers: Values greater than 1 enhance a channel, while values less than 1 reduce it

• Clipping: Use np.clip() to ensure pixel values stay within [0, 1] range

• RGB Conversion: color.gray2rgb() converts single-channel grayscale to 3-channel RGB

• Tint Effects: Warm tints use higher red/yellow values, cool tints use higher blue values

Conclusion

Adding tints to grayscale images in scikit-image involves converting to RGB format and applying color multipliers. The np.clip() function ensures valid pixel ranges, while different multiplier combinations create various artistic effects.