Tutorial: Cropping Images with Python Pillow

Cropping is a common image manipulation task where you extract a specific rectangular portion of an image. Using Python's Pillow library, you can easily crop images and customize the region of interest.

In this tutorial, we’ll cover:

1. Installing Pillow

To install Pillow, use pip:

pip install pillow

2. Basic Cropping

The crop() method in Pillow allows you to define a box and extract a region from the image.

Example: Simple Cropping

from PIL import Image

# Open an image
image = Image.open("example.jpg")

# Define the cropping box (left, upper, right, lower)
crop_box = (100, 100, 400, 400)

# Crop the image
cropped_image = image.crop(crop_box)

# Save and display the cropped image
cropped_image.save("cropped_image.jpg")
cropped_image.show()

Explanation

• The box is defined as (left, upper, right, lower) coordinates.
• (100, 100) is the top-left corner of the box, and (400, 400) is the bottom-right corner.

3. Cropping Based on Specific Coordinates

You can dynamically determine the cropping box based on the image size.

Example: Cropping Dynamically

# Get the image dimensions
width, height = image.size

# Crop the central 50% of the image
left = width * 0.25
upper = height * 0.25
right = width * 0.75
lower = height * 0.75

cropped_center = image.crop((left, upper, right, lower))
cropped_center.save("cropped_center.jpg")
cropped_center.show()

Explanation

• The crop box is calculated as a percentage of the image size.
• This method is useful for consistently cropping multiple images.

4. Center Cropping

Center cropping ensures the cropped region is centered in the image.

Example: Center Cropping with Fixed Size

def center_crop(image, crop_width, crop_height):
    width, height = image.size
    
    # Calculate the cropping box
    left = (width - crop_width) // 2
    upper = (height - crop_height) // 2
    right = (width + crop_width) // 2
    lower = (height + crop_height) // 2
    
    return image.crop((left, upper, right, lower))

# Crop the image to a size of 300x300
center_cropped = center_crop(image, 300, 300)
center_cropped.save("center_cropped.jpg")
center_cropped.show()

5. Cropping Using Aspect Ratio

You can crop an image while maintaining a specific aspect ratio (e.g., 16:9, 1:1).

Example: Cropping to a Specific Aspect Ratio

def crop_to_aspect_ratio(image, aspect_ratio_width, aspect_ratio_height):
    width, height = image.size
    target_aspect_ratio = aspect_ratio_width / aspect_ratio_height
    
    # Calculate new dimensions
    current_aspect_ratio = width / height
    
    if current_aspect_ratio > target_aspect_ratio:
        # Wider than target: crop width
        new_width = int(height * target_aspect_ratio)
        left = (width - new_width) // 2
        right = left + new_width
        top, bottom = 0, height
    else:
        # Taller than target: crop height
        new_height = int(width / target_aspect_ratio)
        top = (height - new_height) // 2
        bottom = top + new_height
        left, right = 0, width

    return image.crop((left, top, right, bottom))

# Crop the image to a 16:9 aspect ratio
cropped_aspect_ratio = crop_to_aspect_ratio(image, 16, 9)
cropped_aspect_ratio.save("cropped_aspect_ratio.jpg")
cropped_aspect_ratio.show()

6. Saving Cropped Images

Save cropped images in the desired format using the save() method:

cropped_image.save("output.jpg", "JPEG", quality=95)

7. Batch Cropping Multiple Images

You can automate cropping for multiple images in a directory.

Example: Batch Cropping

import os

def batch_crop(input_folder, output_folder, crop_box):
    os.makedirs(output_folder, exist_ok=True)
    
    for filename in os.listdir(input_folder):
        if filename.endswith((".jpg", ".jpeg", ".png")):
            image = Image.open(os.path.join(input_folder, filename))
            cropped_image = image.crop(crop_box)
            cropped_image.save(os.path.join(output_folder, filename))
            print(f"Cropped and saved: {filename}")

# Crop the top-left 300x300 pixels for all images in a folder
batch_crop("input_images", "output_images", (0, 0, 300, 300))

8. Practical Examples

8.1 Cropping a Region for Face Detection

# Suppose a face is detected at coordinates (x=50, y=50) with a width and height of 150
face_box = (50, 50, 200, 200)
face_crop = image.crop(face_box)
face_crop.save("face_cropped.jpg")
face_crop.show()

8.2 Creating Thumbnails from Cropped Regions

thumbnail_crop = center_crop(image, 400, 400)
thumbnail_crop.thumbnail((100, 100))
thumbnail_crop.save("thumbnail.jpg")
thumbnail_crop.show()

8.3 Combining Cropping and Resizing

You can crop an image to focus on a specific region and then resize it.

def crop_and_resize(image, crop_box, target_size):
    cropped = image.crop(crop_box)
    resized = cropped.resize(target_size, Image.Resampling.LANCZOS)
    return resized

# Crop a region and resize it to 200x200
processed_image = crop_and_resize(image, (50, 50, 300, 300), (200, 200))
processed_image.save("cropped_and_resized.jpg")
processed_image.show()

9. Summary

Key Methods

• crop(box): Crops the image using the specified box (left, upper, right, lower).
• thumbnail(size): Resizes the cropped region while maintaining the aspect ratio.
• save(): Saves the cropped image in the desired format.

Use Cases

• Extracting specific regions of interest (e.g., faces or objects).
• Preparing images for display in specific aspect ratios.
• Automating tasks with batch cropping.