关于 Pixel Perfect / About Pixel Perfect
Pixel Perfect 是一款免费在线像素画处理工具，专为像素艺术创作者设计。只需上传图片，工具会自动完成网格检测、降噪、色彩量化、重采样等全部处理流程，一键输出高质量像素画。
Pixel Perfect is a free online pixel art processing tool designed for pixel art creators. Simply upload an image and the tool automatically performs grid detection, denoising, color quantization, resampling, and palette extraction — producing high-quality pixel art in one click.
核心功能 / Core Features
- 像素画网格自动检测 (Auto Grid Detection)：智能识别像素艺术中的网格结构，支持不规则网格。Intelligently detects grid structures in pixel art, including irregular grids.
- 智能降噪 (Smart Denoising)：去除图片压缩产生的噪点和伪影，还原干净的像素画效果。Removes compression artifacts and noise to restore clean pixel art.
- OKLAB 色彩空间重采样 (OKLAB Resampling)：使用感知均匀的 OKLAB 色彩空间进行重采样，确保颜色过渡自然。Perceptually uniform color space resampling for natural color transitions.
- 色彩量化 (Color Quantization)：智能合并相近颜色，减少颜色数量同时保持视觉质量。Intelligently merges similar colors, reducing palette size while preserving visual quality.
- 调色板提取 (Palette Extraction)：自动提取 Top 100 常见颜色，支持导出 JSON 和 GPL（GIMP/Aseprite 兼容）格式。Extracts top 100 colors with JSON and GPL (GIMP/Aseprite compatible) export.
- 多种导出方式 (Export Options)：支持导出处理后的完整图片，以及每格 1 像素的真实尺寸图片。Export full-resolution processed image or 1:1 true-size pixel art.
使用场景 / Use Cases
- 从截图或扫描图中提取干净的像素画 / Extract clean pixel art from screenshots or scanned images
- 优化像素画的颜色数量和质量 / Optimize pixel art color count and quality
- 提取像素画的调色板用于其他创作 / Extract palettes from pixel art for use in other projects
- 将放大的像素画还原为 1:1 真实尺寸 / Downscale enlarged pixel art back to true 1:1 size
标准版 vs 专业版 / Standard vs Pro
标准版提供一键全自动处理，适合快速处理。如需逐步调整参数，请使用专业版。
The standard mode provides fully automatic one-click processing. For step-by-step parameter control, use the Pro version.
技术特点 / Technical Highlights
使用自研算法进行网格检测，结合 OKLAB 感知均匀色彩空间进行高质量重采样和色彩量化，实现专业级像素画优化处理。支持 PNG、JPG、WebP、BMP 等常见图片格式。
Uses proprietary algorithms for grid detection, combined with the perceptually uniform OKLAB color space for high-quality resampling and color quantization. Supports PNG, JPG, WebP, BMP and other common image formats.