Face Detection

The face detection feature automatically detects people’s faces in images and makes them the center of the crop, ensuring important facial features remain visible after image transformations.

Face Crop Parameter

Parameter: face_crop
Format 1: Boolean
Format 2: width,height

How It Works

The face crop feature uses intelligent detection algorithms to identify faces in your images and automatically adjusts the crop to keep faces centered and visible.

Boolean Format (Format 1)

When set to true, the optimizer will:

Find all faces it can detect in the image
Ensure that all detected faces remain in frame
Add appropriate padding around the faces
Automatically determine the best crop area


?face_crop=true

Dimension Format (Format 2)

When you specify width and height dimensions, the optimizer will:

Detect all faces in the image
Focus on the center point between all faces (or center of one face if only one is found)
Crop the image to match the specified width and height
Use the detected faces as the center point for the crop


?face_crop=800,1000

Examples

Automatic Face-Centered Crop


# Automatically crop to include all faces with padding
https://cdn.filezen.dev/image.jpg?face_crop=true

Fixed Dimensions with Face Focus


# Crop to 800x1000 pixels centered on detected faces
https://cdn.filezen.dev/image.jpg?face_crop=800,1000

# Square crop centered on faces
https://cdn.filezen.dev/image.jpg?face_crop=500,500

Use Cases

Profile Pictures

Perfect for automatically cropping profile pictures to focus on the person’s face:


?face_crop=400,400

Thumbnail Generation

Generate thumbnails that always keep faces visible:


?face_crop=true&width=200&height=200

Create social media-ready images with proper face positioning:


?face_crop=1200,630

Combining with Other Parameters

Face detection works well with other image manipulation parameters:


# Face-centered crop with quality adjustment
?face_crop=true&quality=80

# Face-centered crop with specific dimensions and blur background
?face_crop=600,600&blur=5

# Face-centered crop with brightness adjustment
?face_crop=true&brightness=10

Best Practices

Use boolean format when you want automatic sizing with optimal face inclusion
Use dimension format when you need specific output dimensions for your application
Combine with quality settings to optimize file size while maintaining face clarity
Test with various face counts to ensure good results with single faces and group photos
Consider padding - the automatic padding ensures faces aren’t cropped too tightly

Limitations

Face detection works best with clear, well-lit faces
Very small faces in large images may not be detected reliably
Heavily obscured or profile faces may not be detected
The feature focuses on human faces and may not detect animal faces