boolean

Perform bitwise boolean operations between two images for masking, compositing, and change detection.

Usage

Description

The boolean command performs bitwise boolean operations between two images at the pixel level. It combines images using AND, OR, or XOR (EOR) logic operations. This is useful for masking, pattern detection, image differencing, and creative compositing effects.

Options

Operations

AND Operation

Bitwise AND - Keeps pixels present in both images.

Logic: output = input AND operand

Effect:

• Black pixel (0) + Any pixel = Black (0)
• White pixel (255) + White pixel (255) = White (255)
• Removes areas where operand is black
• Preserves areas where both are bright

Use cases:

• Apply masks to extract regions
• Remove backgrounds
• Intersection of two images
• Selective area extraction

OR Operation

Bitwise OR - Combines pixels from both images.

Logic: output = input OR operand

Effect:

• Black pixel (0) + Black pixel (0) = Black (0)
• Any bright pixel results in bright output
• Adds brightness from both images
• Combines patterns

Use cases:

• Merge patterns
• Combine multiple exposures
• Add watermarks/overlays
• Union of two images

EOR Operation (XOR)

Exclusive OR - Highlights differences between images.

Logic: output = input XOR operand

Effect:

• Same pixels (0,0 or 255,255) = Black (0)
• Different pixels = Bright
• Shows where images differ
• Creates pattern effects

Use cases:

• Change detection
• Find differences between versions
• Create symmetric patterns
• Image comparison

Examples

Masking with AND

# Apply black/white mask to extract region
mediaproc image boolean photo.jpg --operation and --operand mask.png -o masked.jpg

# Extract only masked area
mediaproc image boolean image.png --operation and --operand alpha-mask.png

# Remove background using inverse mask
mediaproc image boolean portrait.jpg --operation and --operand foreground-mask.png -o no-bg.jpg

Combining with OR

# Combine two patterns
mediaproc image boolean pattern1.png --operation or --operand pattern2.png -o combined.png

# Add watermark overlay
mediaproc image boolean photo.jpg --operation or --operand watermark.png -o marked.jpg

# Merge light sources
mediaproc image boolean base.jpg --operation or --operand light-effect.png -o bright.jpg

Change Detection with XOR

# Find differences between two versions
mediaproc image boolean before.jpg --operation eor --operand after.jpg -o diff.jpg

# Detect changes in scene
mediaproc image boolean frame1.png --operation eor --operand frame2.png -o changes.png

# Compare image versions
mediaproc image boolean original.jpg --operation eor --operand edited.jpg -o differences.jpg

Creative Effects

# Create symmetric patterns
mediaproc image boolean texture.png --operation eor --operand flipped-texture.png -o pattern.png

# Artistic compositing
mediaproc image boolean photo.jpg --operation or --operand artistic-overlay.png -o art.jpg

# Pattern generation
mediaproc image boolean grid1.png --operation eor --operand grid2.png -o geometric.png

Batch Processing

# Apply same mask to multiple images
mediaproc image boolean "photos/*.jpg" --operation and --operand common-mask.png

# Combine each with overlay
mediaproc image boolean "images/*.png" --operation or --operand overlay.png -v

Requirements

Image Dimensions

Important: Both input and operand images should ideally have the same dimensions.

• Mismatched sizes may cause cropping or errors
• Sharp.js will try to handle size differences
• Best results with exact dimension matches

Format Compatibility

Works with all image formats:

• JPEG/JPG: Good for photos
• PNG: Best for masks (transparency)
• WebP: Modern format support
• GIF, TIFF: Also supported

Color Depth

• Operations work on all channels (R, G, B, Alpha)
• Results are more predictable with grayscale
• Consider converting to grayscale first for clarity

Practical Use Cases

Masking and Extraction

Remove Background:

# Create white background where subject is, black elsewhere
# Then use AND to keep only subject
mediaproc image boolean photo.jpg --operation and --operand subject-mask.png

Extract Region:

# Use white mask for area of interest
mediaproc image boolean image.jpg --operation and --operand roi-mask.png -o extracted.jpg

Pattern Creation

Symmetric Designs:

# XOR with flipped version creates symmetry
mediaproc image flip texture.png -o flipped.png
mediaproc image boolean texture.png --operation eor --operand flipped.png -o symmetric.png

Change Detection

Security Monitoring:

# Detect changes in surveillance images
mediaproc image boolean cam-before.jpg --operation eor --operand cam-after.jpg -o motion-detected.jpg

Version Comparison:

# See what changed between edits
mediaproc image boolean draft1.png --operation eor --operand draft2.png -o changes.png

Watermarking

Invisible Watermark (Steganography):

# Embed pattern using OR
mediaproc image boolean original.jpg --operation or --operand subtle-watermark.png -o protected.jpg

Operation Examples

AND: Masking

Input 1 (photo.jpg):

Operand (mask.png):

Result:

OR: Combining

Input 1 (base.jpg):

Operand (pattern.png):

Result:

XOR: Difference

Input 1 (before.jpg):

Operand (after.jpg):

Result:

Best Practices

1. Match Dimensions: Ensure both images are same size
2. Use Grayscale Masks: Simpler and more predictable
3. Test First: Use --dry-run before batch operations
4. High Quality: Use -q 95+ to avoid compression artifacts
5. PNG for Masks: Use PNG format for masks (supports transparency)
6. Convert First: Consider converting both to same format first

Performance Tips

• Operand is loaded once and reused for batch operations
• Works quickly even on large images
• Minimal memory overhead
• Fast bitwise operations at hardware level

Common Patterns

Create Mask from Image:

# Convert to grayscale, threshold, then use as mask
mediaproc image grayscale photo.jpg -o gray.jpg
mediaproc image threshold gray.jpg --level 128 -o mask.png
mediaproc image boolean original.jpg --operation and --operand mask.png

Combine Multiple Overlays:

# OR multiple patterns together
mediaproc image boolean base.jpg --operation or --operand overlay1.png -o temp.jpg
mediaproc image boolean temp.jpg --operation or --operand overlay2.png -o final.jpg

Detect Motion:

# XOR consecutive frames
mediaproc image boolean frame-t0.jpg --operation eor --operand frame-t1.jpg -o motion.jpg

Troubleshooting

Dimension mismatch error:

• Resize both images to same dimensions first
• Use mediaproc image resize to match sizes

Unexpected all-black output:

• Check if operand mask is inverted (black/white swapped)
• Verify AND operation is appropriate for use case

Unexpected all-white output:

• Check if using OR with bright images
• May need to use AND or XOR instead

Results not visible:

• Try adjusting brightness/contrast of output
• XOR results may be very dark - use normalize command

Operation not recognized:

• Use lowercase: and, or, eor (not AND, OR, XOR)
• XOR is called eor in this command

Advanced Techniques

Multi-step Masking

# Complex mask creation
mediaproc image boolean img.jpg --operation and --operand mask1.png -o step1.jpg
mediaproc image boolean step1.jpg --operation or --operand mask2.png -o final.jpg

Pattern Generation

# Create kaleidoscope effect with XOR
mediaproc image rotate pattern.png --angle 90 -o rot90.png
mediaproc image boolean pattern.png --operation eor --operand rot90.png -o kaleidoscope.png

See Also

• composite - Layer images with blend modes
• extract - Extract channels or regions
• threshold - Create binary masks
• grayscale - Convert for simpler boolean operations
• normalize - Adjust boolean operation results