p2ascii

p2ascii is a Python-based tool that converts images into ASCII art. It supports:

• Simple mode: brightness-based mapping of pixels to ASCII characters.
• Complex mode: edge-aware rendering using Sobel gradients to compute orientation and magnitude.

It can output:

• ASCII text, optionally with ANSI color codes.
• ASCII images, where characters are rendered as image blocks.

Note: The output that is produced is the same dimension has the original both in image and in text. So using it in text keep in mind small images are better.

• Sobel-based edge detection for orientation-aware rendering.
• Optional color support for both text and image modes.
• Configurable edge sensitivity (<thresh>).
• Terminal or image file output.
• Simple and complex conversion modes.
• Mode with transparency

You can install p2ascii directly from the AUR using your favorite AUR helper:

yay -S p2ascii

Or manually:

git clone https://aur.archlinux.org/p2ascii.git
cd p2ascii
makepkg -si

• Python 3.7+
• opencv-python
• numpy

• sc2image <img> [--transparent] Convert image to ASCII image
• sc2text <img> Convert image to ASCII text (stdout)
• sc2cimage <img> [--transparent] Colored ASCII image
• sc2ctext <img> Colored ASCII text (stdout)

• cc2image <img> <thresh> [--transparent] ASCII image using edge orientation
• cc2text <img> <thresh> ASCII text with edge symbols (stdout)
• cc2cimage <img> <thresh> [--transparent] Colored ASCII image with edge awareness
• cc2ctext <img> <thresh> Colored ASCII text with edge symbols

Used in complex conversion to control edge sensitivity based on gradient magnitude.

• auto: Automatically computes threshold as the 90th percentile of non-zero magnitudes:
• 0–255: Manual threshold.
  • Higher = fewer edges (only strong ones)
  • Lower = more edges (fine details)

When included, the output ASCII image will have a transparent background, displaying only the characters. This option is applicable to image output commands: sc2image, sc2cimage, cc2image, and cc2cimage.

Motivation

My motivation for this project stems from a captivating video by Acerola, which you can find here: https://www.youtube.com/watch?v=gg40RWiaHRY. Inspired by his work, I wanted to challenge and test my skills in computer vision.