img2ascii-py

A high-performance Python library and command-line (CLI) utility that converts images into beautiful, styled ASCII art or pixel-exact HTML/CSS output.

Built with performance in mind using fully vectorized NumPy operations, img2ascii-py generates highly optimized terminal ANSI truecolor prints and run-length compressed HTML web pages.

---

Features

• Vectorized Core: Color pre-processing (gamma, brightness, contrast) and block downsampling are fully vectorized using NumPy.
• Terminal Truecolor: Output ASCII art in 24-bit ANSI colors with a state-machine that minimizes escape-sequence overhead.
• Pixel-Exact HTML: Render pixel-art pages utilizing Run-Length Encoding (RLE) to bundle matching color spans and optimize filesize. Supports both resize (pre-sampling aspect scaling) and css (visual aspect scaling via line-height to preserve literal source pixel data) modes.
• Pixel-Exact SVG Vector Output: Convert images into scalable, clean SVG vector graphics with RLE-optimized <rect> nodes.
• Curated Preset Ramps: Includes preset charsets:
  • standard: .:-=+*#%@
  • detailed: $@B%8&WM#*oahkbdpqwmZO0QLCJUYXzcvunxrjft/\|()1{}[]?-_+~<>i! lI;:,"^'. `
  • blocks: ░▒▓█
  • binary: #
  • minimal: .o0@
  • Custom: Pass any string directly as your ramp!
• Smart Image Handling: Automatically transposes images based on EXIF rotation tags and handles transparent PNG alpha channels gracefully.
• Full Stream Piping: Pipe binary image streams directly into the CLI via stdin.

---

Installation

Install the package directly from PyPI:

pip install img2ascii-py

Optional Extras

• To enable faster image sampling (via Numba JIT compilation):

  pip install "img2ascii-py[fast]"
  

• To enable edge-detection enhancement filters:

  pip install "img2ascii-py[edges]"
  

• Install all features at once:

  pip install "img2ascii-py[fast,edges]"
  

---

CLI Usage

When installed, the img2ascii command is added to your path.

img2ascii --help

Quick Examples

1. Basic Grayscale ASCII Art

Scale an image to a custom width and save the text file:

img2ascii path/to/image.jpg --width 80 > art.txt

2. Colored Terminal Print

Display the image directly inside the terminal with 24-bit ANSI colors:

img2ascii path/to/image.jpg --width 100 --color

3. Pixel-Exact HTML/CSS Output

Convert an image to a pixel-perfect HTML webpage using color-grouped HTML spans:

img2ascii path/to/image.jpg --width 150 --mode pixel -o page.html

4. SVG Vector Graphic Output

Convert an image to a scalable SVG vector graphic file:

img2ascii path/to/image.jpg --width 120 --mode svg -o graphic.svg

5. CSS Aspect Ratio Mode

Save a pixel-exact HTML file where visual aspect-ratio correction is done entirely in CSS (preserving literal original pixel coordinates):

img2ascii path/to/image.jpg --width 100 --mode pixel --aspect-mode css -o page.html

6. Piping from Standard Input

Send binary stream output into the converter:

cat input.png | img2ascii - --width 60 -o output.txt

7. JIT-Accelerated Fast Mode

Enable performance JIT acceleration using Numba for processing high-resolution files:

img2ascii path/to/large_image.jpg --width 200 --fast -o output.txt

8. Edge Detection Enhancement

Overlay edge boundaries using Sobel filters to construct line art matching visual structures:

img2ascii path/to/line_art.png --width 100 --edges -o output.txt

9. Palette Quantization (Color Compression)

Trade a small amount of color fidelity to yield significantly smaller file sizes with better run-length compression (pixel/SVG modes only):

img2ascii path/to/image.jpg --width 100 --mode pixel --palette-size 16 -o quantized.html

10. HTML Export of ASCII Art

Generate a colored, fully-styled HTML webpage that uses custom mapped ASCII glyphs:

img2ascii path/to/image.jpg --width 100 --mode html -o ascii_art.html

---

Library API Reference

You can also import and use img2ascii programmatically in your Python scripts.

Grayscale or Colored ASCII Art

from img2ascii.api import convert_to_ascii, AsciiConfig

# Configure settings
config = AsciiConfig(
    width=80,
    char_aspect=2.0,       # Adjusts height/width ratio for terminal fonts
    charset="standard",    # Supports presets: standard, detailed, blocks, binary, minimal
    color=True,            # Enable ANSI color escape codes
    auto_contrast=True,    # Stretch luma values for maximum dynamic range
    fast=True,             # Optional: Enable Numba JIT acceleration
    edges=True             # Optional: Enable SciPy Sobel edge-enhancement
)

# Render from file path, raw bytes, or a PIL Image object
art = convert_to_ascii("image.jpg", config)
print(art)

Pixel-Exact Web Output (HTML)

from img2ascii.api import convert_to_pixels, PixelConfig

config = PixelConfig(
    width=120,
    bg_color="#111111",
    glyph="█",             # Character block used to draw each pixel
    aspect_mode="css",     # aspect correction via CSS line-height (keeps raw pixel resolution)
    fast=True,             # Optional: Enable Numba JIT acceleration
    palette_size=16        # Optional: Quantize to 16 colors for compression
)

html_code = convert_to_pixels("image.png", config)
with open("output.html", "w") as f:
    f.write(html_code)

Pixel-Exact Vector Output (SVG)

from img2ascii.api import convert_to_svg, PixelConfig

config = PixelConfig(
    width=120,
    bg_color="#111111",
    fast=True,             # Optional: Enable Numba JIT acceleration
    palette_size=16        # Optional: Quantize to 16 colors for compression
)

svg_code = convert_to_svg("image.png", config)
with open("output.svg", "w") as f:
    f.write(svg_code)

HTML Export of ASCII Art (HTML ASCII)

from img2ascii.api import convert_to_html, HtmlConfig

config = HtmlConfig(
    width=120,
    charset="standard",
    bg_color="#111111",
    aspect_mode="resize",  # resize or css
    fast=True,             # Optional: Enable Numba JIT acceleration
    edges=True,            # Optional: Enable SciPy Sobel edge-enhancement
    palette_size=16        # Optional: Quantize to 16 colors for compression
)

html_code = convert_to_html("image.png", config)
with open("output.html", "w") as f:
    f.write(html_code)

---

License

This project is open-source and licensed under the MIT License.