How Character Density Mapping Works

The Algorithm Behind Image-to-ASCII Conversion

Character density mapping is the core technique used to convert images into ASCII art. The idea is simple: each character occupies the same space in a monospace font, but different characters fill different amounts of that space with ink (or lit pixels). A character like @ fills most of its cell, while a period . leaves most of it empty.

Step 1: Divide the Image into Cells

The image is divided into a grid of rectangular cells. Each cell corresponds to one character in the output. Because monospace characters are roughly twice as tall as they are wide, each cell has a 1:2 width-to-height ratio to prevent the output from appearing stretched.

Step 2: Calculate Average Brightness

For each cell, the algorithm calculates the average brightness of all pixels within it. The standard formula uses the ITU-R BT.709 luminance coefficients:

brightness = 0.2126 * R + 0.7152 * G + 0.0722 * B

These weights reflect how the human eye perceives brightness — green contributes the most, followed by red, then blue. This produces perceptually accurate results compared to a simple average of the three channels.

Step 3: Map Brightness to Characters

The brightness value (0-255, normalized to 0-1) is mapped to an index in the character density scale. In a scale like @#%*+=-:. , the darkest character (@) maps to brightness 0 and the lightest (space) maps to brightness 1.

charIndex = floor(brightness * (charSetLength - 1))
character = charSet[charIndex]

Inversion

On dark terminal backgrounds, you typically want bright image regions to map to sparse characters (which show the background color) and dark regions to map to dense characters. On light backgrounds, you may want to invert this mapping.

Color Extension

For colored ASCII art, the algorithm additionally stores the average RGB color of each cell and wraps the output character in an HTML <span> with an inline color style. This preserves the original color information alongside the brightness-based character selection.