gitmyhub

Pixel-Fixer

Python ★ 1 updated 13d ago

A lightweight Vibe-coded tool designed to take AI generated pixel art Images and align pixels to grid and reduce colour pallette.

A drag-and-drop desktop tool that fixes AI-generated pixel art by snapping pixels back to a clean grid and reducing the color palette using a GIMP palette file or automatic quantization.

Pythonsetup: easycomplexity 1/5

Pixel Fixer is a small desktop tool that cleans up pixel art images generated by AI tools. When an AI produces pixel art, the result often has pixels that do not align to a clean grid, blurry scaling artifacts, colors that drift slightly from the intended palette, and more colors than traditional pixel art uses. This tool fixes those problems.

You drag an image into the program's window. Using a signal processing technique called Fast Fourier Transform, it detects the underlying pixel grid in the image and snaps every pixel back to that grid. For colors, you have two options. You can load a standard palette file from the GIMP image editor (files with the .gpl extension) by placing them in a palettes folder, and the tool will map every pixel to the nearest matching color in that palette. Alternatively, you can switch to an automatic mode where the tool builds an optimized palette directly from the image itself and reduces it to a number of colors you choose.

The example in the README shows an image reduced from 256 colors down to 16 using this process. The tool includes several example outputs with different palette styles applied to the same image.

Installation is a download and extract from the GitHub releases page. You then populate a palettes folder with any .gpl files you want to use. No license is mentioned in the README.

Where it fits