Basic Image Editing

Image manipulation tool supporting transformations, format conversions, transparency operations, and optimization. Useful for resizing images, removing backgrounds, converting formats, reducing file sizes, and preparing images for web or applications.

Quick Start

bash

uv run scripts/image_edit.py INPUT -o OUTPUT [options]

Operations

Image Info

bash

uv run scripts/image_edit.py input.png --info

Displays: dimensions, format, file size, color mode, DPI.

Rotate

bash

uv run scripts/image_edit.py input.png -o output.png --rotate 90
uv run scripts/image_edit.py input.png -o output.png --rotate 45

Flip

bash

uv run scripts/image_edit.py input.png -o output.png --flip horizontal
uv run scripts/image_edit.py input.png -o output.png --flip vertical

Resize

bash

# Exact dimensions
uv run scripts/image_edit.py input.png -o output.png --width 800 --height 600

# Width only (maintains aspect ratio)
uv run scripts/image_edit.py input.png -o output.png --width 800

# Height only (maintains aspect ratio)
uv run scripts/image_edit.py input.png -o output.png --height 600

Format Conversion

Convert by specifying output extension:

bash

uv run scripts/image_edit.py input.png -o output.webp
uv run scripts/image_edit.py input.heic -o output.jpg

Supported: JPEG, PNG, WebP, TIFF, HEIC/HEIF.

Grayscale

bash

uv run scripts/image_edit.py input.png -o output.png --grayscale

Converts to single-channel grayscale (mode 'L').

Transparency

bash

# Remove alpha (white background)
uv run scripts/image_edit.py input.png -o output.jpg --remove-transparency

# Replace with color
uv run scripts/image_edit.py input.png -o output.png --replace-transparency red
uv run scripts/image_edit.py input.png -o output.png --replace-transparency "#FF5500"

If image has no alpha channel, prints a note and continues.

Extract Alpha Mask

bash

uv run scripts/image_edit.py input.png -o mask.png --extract-mask

Extracts the alpha channel as a 1-channel grayscale mask. Requires RGBA image.

Alpha Blend (Apply Mask)

bash

uv run scripts/image_edit.py input.png -o output.png --mask mask.png

Applies a grayscale mask as the alpha channel, creating an RGBA composite. Mask is automatically resized to match image dimensions if needed.

Auto-Crop Transparency

bash

uv run scripts/image_edit.py input.png -o output.png --autocrop-transparency 0
uv run scripts/image_edit.py input.png -o output.png --autocrop-transparency 5

Crops transparent borders from image. Threshold (0-100%) controls which pixels are considered transparent:

•0 = only fully transparent pixels (alpha=0)
•5 = pixels with alpha < 5% (≈12/255)
•100 = all pixels considered transparent

Padding

bash

# All sides equal
uv run scripts/image_edit.py input.png -o output.png --pad 20

# Vertical, horizontal
uv run scripts/image_edit.py input.png -o output.png --pad 10,20

# Top, right, bottom, left
uv run scripts/image_edit.py input.png -o output.png --pad 10,20,30,40

# With color
uv run scripts/image_edit.py input.png -o output.png --pad 20 --pad-color blue

# Edge pixel replication
uv run scripts/image_edit.py input.png -o output.png --pad 20 --pad-edge

Cropping

bash

# All sides equal
uv run scripts/image_edit.py input.png -o output.png --crop 50

# Top, right, bottom, left
uv run scripts/image_edit.py input.png -o output.png --crop 10,20,30,40

File Size Reduction

bash

uv run scripts/image_edit.py input.png -o output.jpg --max-size 1
uv run scripts/image_edit.py input.png -o output.webp --max-size 0.5

Uses binary search to find optimal quality/dimensions.

Color Formats

Format	Examples
Named	`red`, `blue`, `coral`, `darkslategray` (140+ CSS colors)
Hex	`#RGB`, `#RRGGBB`, `#RRGGBBAA`
RGB	`255,128,0` or `rgb(255,128,0)`
HSL	`hsl(120,100%,50%)`

Combining Operations

Operations apply in order: rotate → flip → autocrop → crop → resize → pad → mask → grayscale → transparency.

bash

uv run scripts/image_edit.py input.png -o output.jpg \
    --rotate 90 \
    --autocrop-transparency 5 \
    --width 800 \
    --pad 20 --pad-color white

Important Notes

•Output filename: Always use a different output filename than the input to avoid overwriting the original. A good practice is to append a descriptive suffix to the input filename (e.g., image.png → image_rotated.png, photo.jpg → photo_resized.jpg).