ArcGrid 是一款用于生成图标比例绘图和辅助线的 Web 应用原型,可自动为设计作品提取并标注专业的几何制图规范。
✨ 在线体验: https://charlex.me/lab/arcgrid/
- 🔄 双轨输入系统 - 支持直接粘贴/上传 SVG,或通过 Gemini AI 将图像 (PNG/JPG) 一键智能矢量化
- 🧠 智能几何分析 - 自动识别图形特征,生成精确的圆、圆弧和对齐线计算方案
- 🛠️ 强悍的 SVG 解析 - 自动转换描边为填充、过滤冗余元素并完美应对复杂的变换矩阵
- 🎨 精细化交互体验 - 实时动态调整检测容差,支持图层切换以及辅线颜色、粗细的自定义
- ✨ 动感可视化 - 渲染过程自带优美的辅助线生长动画
- 📥 高质量规格导出 - 一键将生成的制图规范及网格导出为 SVG 或 PDF 格式
# 克隆仓库并安装依赖
git clone <repository-url>
cd ArcGrid
npm install
# 启动开发服务器 (自动运行 Tailwind & Wrangler)
npm run dev访问
http://localhost:8788即可开始使用。
完全免环境配置。直接在网页设置面板输入 Gemini API Key 即可启用 AI 视觉功能:
- 原生支持 gemini-3.1-flash-image-preview 及 gemini-2.5-flash-image
- 密钥安全存储于浏览器本地 (Local Storage)
- (注:仅在使用“图像转 SVG”实验性功能时需要配置)
- 前端 - 原生 HTML5 / Tailwind CSS / ES Modules
- 后端 - Cloudflare Pages Functions (Hono.js)
- 大语言模型 - Google Gemini API
ArcGrid is a specialized web application prototype that automates the generation of professional construction lines and geometric annotations for logo designs.
✨ Live Demo: https://charlex.me/lab/arcgrid/
- 🔄 Double-Track Input - Directly paste/upload SVGs, or use Gemini AI to automatically vectorize raster images (PNG/JPG).
- 🧠 Intelligent Analysis - Automatically detects circles, arcs, and alignment lines to create precise construction candidates.
- 🛠️ Robust SVG Processing - Handles stroke-to-fill conversions, ignores non-renderable
<defs>, and tackles complex transformations. - 🎨 Interactive UI - Real-time tolerance adjustment, visibility toggles for layers, and customizable guide colors/weights.
- ✨ Stunning Visuals - Features an engaging animated line-drawing effect during validation and rendering.
- 📥 Professional Export - Export your customized construction guides as high-quality SVGs or PDFs natively.
# Clone the repo and install dependencies
git clone <repository-url>
cd ArcGrid
npm install
# Start the development server
npm run devThe application will be running at
http://localhost:8788.
No .env required. Input your Gemini API Key directly in the web settings panel:
- Native support for gemini-3.1-flash-image-preview and gemini-2.5-flash-image.
- Keys are securely stored in your browser's local storage.
- (Note: Gemini API is only required for the experimental Image-to-SVG vectorization feature).
- Frontend - Vanilla HTML5 / Tailwind CSS / ES Modules
- Backend / Hosting - Cloudflare Pages Functions (Hono.js)
- AI Integration - Google Gemini API
