brew install imagemagick

convert input.png -dither Riemersma output.png
// or 
convert input.png -dither FloydSteinberg output.png

convert input.png -monochrome -dither Riemersma output.png
// or 
convert input.png -monochrome -dither FloydSteinberg output.png

When using the -posterize option you need to add a level. Low numbers like 2, 3, 4, 5 have the most effect. → Official Reference

convert input.png -posterize 2 -dither FloydSteinberg output.png

Reduce the number of colors in an image to the colors used by another image (e.g. remapImage.png). → Official Reference

convert input.png -remap remapImage.png -dither FloydSteinberg output.png

The remap (input) image

The output image

Set the preferred number of colors in the image. → Official Reference

convert input.png -colors 4 -dither FloydSteinberg output.png

2 Colors

5 Colors

10 Colors

Use -ordered-dither to apply threshold mapped dither patterns, using uniform color maps, rather than specific color maps. → Official Reference.

convert input.png -colors 10 -ordered-dither o2x2 -dither FloydSteinberg output.png

-ordered-dither checks

-ordered-dither h4x4a

-ordered-dither h8x8a

-ordered-dither o2x2

Use convert list -threshold to see all available thresholds.

convert input.png -colors 10 -morphology Dilate Octagon:10 -dither FloydSteinberg output.jpg

The 'Hit-And-Miss' morphology method, also commonly known as "HMT" in computer science literature, is a high level morphology method that is specifically designed to find and locate specific patterns in images. It does this by looking for a specific configuration of 'foreground' and 'background' pixels around the 'origin'. → Official Reference.

convert input.png -colors 10 -morphology Hit-and-Miss '2x1:1,0' -dither FloydSteinberg output.jpg
// or 
convert input.png -colors 10 -morphology Hit-and-Miss '3x1:1,-,0' -dither FloydSteinberg output.jpg

This method subtracts them from the original image. → Official Reference.

convert input.png -colors 10 -morphology Thinning '10x1+1+0:1,1,1,1,1,1,1,1,1,1' output.jpg

Create a Skeleton from your image. This works best when you make your images monochrome first. → Official Reference.

convert input.png -monochrome -morphology Thinning:-1 Skeleton output.jpg

• CLI for Artists + Designers
• Official Docs