AI 驱动的智能抠图工具 | AI-Powered Background Remover
🚀 秒级处理 · 🔒 本地运行 · 💯 完全免费 · ✨ 高质量输出
RemoveBG PicGo 是一款基于 AI 技术的在线智能抠图工具。使用先进的 RMBG-1.4 和 U2Net 深度学习模型，可以精准识别图片主体并智能去除背景。
🌟 核心亮点：所有处理完全在浏览器本地进行，图片不会上传到任何服务器，充分保护您的隐私！
访问 rmbg.picgo.studio 立即体验！
| 功能 | 描述 |
|---|---|
| 🎯 精准抠图 | 使用先进的 AI 模型，精准识别主体并去除背景 |
| 🔄 模型切换 | 支持 RMBG-1.4 和 U2Net 两种模型，可根据场景选择最佳效果 |
| 🧵 Web Worker 后台处理 | 模型推理在独立线程运行，主线程完全不阻塞，处理期间可自由操作 |
| 🔒 隐私安全 | 所有处理完全在本地进行，图片不会上传到任何服务器 |
| 👁️ 实时对比 | 处理完成后可以左右对比查看处理前后的效果差异 |
| 🎨 Neo-Brutalism 设计 | 采用现代 Neo-Brutalism 设计风格，提供独特的视觉体验 |
| 📥 高清导出 | 支持导出原始分辨率的透明背景图片，保持图片质量 |
| 🌍 多语言支持 | 支持简体中文、English、日本語、한국어、Русский |
| 🎭 主题切换 | 支持浅色和深色模式，可跟随系统设置自动切换 |
| 📱 PWA 支持 | 可安装为桌面应用，支持离线使用 |
- 特点: 通用场景效果更好，适合大多数图片，推荐优先使用
- 适用场景: 人像、产品、动物等各类主体抠图
- 特点: 某些特定场景下可能效果更佳，可作为备选方案
- 适用场景: 复杂背景、细节丰富的图片
- 上传图片 - 拖拽或点击上传需要处理的图片（支持 JPG、PNG、WebP）
- 自动处理 - AI 模型自动识别主体并去除背景
- 对比查看 - 使用对比滑块查看处理前后效果
- 调整视图 - 使用缩放和平移功能查看细节
- 切换模型 - 如果效果不理想，可在设置中切换其他模型重新处理
- 下载结果 - 预览效果满意后下载透明背景图片
| 类别 | 技术 |
|---|---|
| 框架 | Next.js 16 + React 19 |
| 语言 | TypeScript 5 |
| 样式 | Tailwind CSS 4 |
| AI 推理 | ONNX Runtime Web + Web Worker |
| AI 模型 | RMBG-1.4 + U2Net |
| 动画 | Framer Motion |
| 国际化 | next-intl |
| UI 组件 | Radix UI |
# 克隆项目
git clone https://github.com/isboyjc/removebg.git
cd removebg
# 安装依赖
pnpm install
# 启动开发服务器
pnpm dev
# 构建生产版本
pnpm build
# 启动生产服务器
pnpm start访问 http://localhost:3000 查看应用。
RemoveBG PicGo 支持以下语言：
- 🇨🇳 简体中文
- 🇺🇸 English
- 🇯🇵 日本語
- 🇰🇷 한국어
- 🇷🇺 Русский
支持浅色模式和深色模式，可跟随系统设置自动切换。
- Chrome/Edge: >= 90
- Firefox: >= 88
- Safari: >= 14
- 需要支持 WebAssembly 和 Cache API
RemoveBG PicGo is an AI-powered online background removal tool. Using advanced RMBG-1.4 and U2Net deep learning models, it can accurately identify subjects and intelligently remove backgrounds.
🌟 Key Highlight: All processing is done entirely in your browser locally. Images are never uploaded to any server, fully protecting your privacy!
Visit rmbg.picgo.studio to try it now!
| Feature | Description |
|---|---|
| 🎯 Precise Removal | Uses advanced AI models to accurately identify subjects and remove backgrounds |
| 🔄 Model Switching | Supports RMBG-1.4 and U2Net models, choose the best for your scenario |
| 🧵 Web Worker Processing | Model inference runs in a separate thread, keeping the main thread responsive during processing |
| 🔒 Privacy Safe | All processing is done locally, images are never uploaded to any server |
| 👁️ Real-time Comparison | Compare before and after effects with a side-by-side slider |
| 🎨 Neo-Brutalism Design | Modern Neo-Brutalism design style for a unique visual experience |
| 📥 HD Export | Export results at original resolution with transparent background, maintaining image quality |
| 🌍 Multi-language | Supports Simplified Chinese, English, Japanese, Korean, Russian |
| 🎭 Theme Switching | Supports light and dark mode, can automatically switch based on system settings |
| 📱 PWA Support | Install as a desktop app with offline support |
- Features: Better performance for general scenarios, suitable for most images, recommended first choice
- Use Cases: Portraits, products, animals, and various subject removal
- Features: May perform better in certain specific scenarios, can be used as an alternative
- Use Cases: Complex backgrounds, detail-rich images
- Upload Image - Drag and drop or click to upload an image (supports JPG, PNG, WebP)
- Auto Processing - AI model automatically identifies subject and removes background
- Compare View - Use the comparison slider to view before and after effects
- Adjust View - Use zoom and pan features to view details
- Switch Model - If the result is not ideal, switch to another model in settings and reprocess
- Download Result - Preview the result and download the transparent background image
| Category | Technology |
|---|---|
| Framework | Next.js 16 + React 19 |
| Language | TypeScript 5 |
| Styling | Tailwind CSS 4 |
| AI Inference | ONNX Runtime Web + Web Worker |
| AI Models | RMBG-1.4 + U2Net |
| Animation | Framer Motion |
| i18n | next-intl |
| UI Components | Radix UI |
# Clone the repository
git clone https://github.com/isboyjc/removebg.git
cd removebg
# Install dependencies
pnpm install
# Start development server
pnpm dev
# Build for production
pnpm build
# Start production server
pnpm startOpen http://localhost:3000 to view the application.
RemoveBG PicGo supports the following languages:
- 🇨🇳 Simplified Chinese
- 🇺🇸 English
- 🇯🇵 Japanese
- 🇰🇷 Korean
- 🇷🇺 Russian
Supports light and dark mode, can automatically switch based on system settings.
- Chrome/Edge: >= 90
- Firefox: >= 88
- Safari: >= 14
- Requires WebAssembly and Cache API support
