Determine the type of an image in Python?

In this section, we are going to see how to determine the type of image file programmatically using Python. Consider a situation where you have hundreds of image files in a directory and want to filter specific image formats like JPEG, PNG, or GIF.

Python provides the imghdr library to determine the type of an image contained in a file or byte stream.

Installation

The imghdr module is a standard library package that comes with Python 3.6 and higher installations. However, if you need to install it separately, run the following command ?

pip install imghdr

To verify the installation, import the module in your Python shell ?

import imghdr
print("imghdr module imported successfully!")

imghdr module imported successfully!

Syntax

The imghdr package defines the following function ?

imghdr.what(filename[, h])

Parameters

• filename: Path to the image file to test. Returns a string describing the image type.

• h (optional): If provided, filename is ignored and h is assumed to contain the byte stream to test.

Supported Image Formats

Below are the image formats recognized by the imghdr package ?

Basic Example

Let's create a simple example to determine image types ?

import imghdr

# Create a sample image data (PNG signature)
png_signature = b'\x89PNG\r\n\x1a\n'

# Test with byte stream
image_type = imghdr.what(None, png_signature)
print(f"Image type from byte stream: {image_type}")

# You can also test with actual files
# image_type = imghdr.what('sample.jpg')
# print(f"Image type: {image_type}")

Image type from byte stream: png

Practical Implementation

Here's a practical function to identify and organize image files by their actual format ?

import imghdr
import os

def identify_image_type(image_path):
    """
    Identify the actual image type regardless of file extension
    """
    try:
        file_type = imghdr.what(image_path)
        if file_type:
            print(f"File: {os.path.basename(image_path)}")
            print(f"Detected type: {file_type}")
            return file_type
        else:
            print(f"Unknown or invalid image format: {image_path}")
            return None
    except Exception as e:
        print(f"Error processing {image_path}: {e}")
        return None

# Example usage
sample_files = ['image.jpg', 'photo.png', 'graphic.gif']

for filename in sample_files:
    print(f"Testing: {filename}")
    # In real scenario, you would pass actual file paths
    print("---")

Testing: image.jpg
---
Testing: photo.png
---
Testing: graphic.gif
---

Filtering Images by Type

Here's how to filter specific image types from a directory ?

import imghdr
import os

def filter_images_by_type(directory, target_type='jpeg'):
    """
    Filter images by their actual type (not extension)
    """
    target_images = []
    
    # Simulate directory listing
    sample_files = ['photo1.jpg', 'image2.png', 'graphic.gif', 'document.pdf']
    
    for filename in sample_files:
        filepath = os.path.join(directory, filename)
        
        # In real implementation, you would check if file exists
        # and use actual file path with imghdr.what(filepath)
        
        # Simulating different file types for demonstration
        if filename.endswith('.jpg'):
            detected_type = 'jpeg'
        elif filename.endswith('.png'):
            detected_type = 'png'
        elif filename.endswith('.gif'):
            detected_type = 'gif'
        else:
            detected_type = None
            
        if detected_type == target_type:
            target_images.append(filename)
            print(f"Found {target_type}: {filename}")
    
    return target_images

# Find all JPEG images
jpeg_files = filter_images_by_type('/path/to/images', 'jpeg')
print(f"\nTotal JPEG files found: {len(jpeg_files)}")

Found jpeg: photo1.jpg
Total JPEG files found: 1

Conclusion

The imghdr module provides a reliable way to determine actual image formats regardless of file extensions. Use imghdr.what() to identify image types programmatically and build robust image processing applications.