Metadata-Version: 2.4
Name: vectify-simplify
Version: 1.0.0
Summary: Enterprise-grade AI image upscaling and multi-threaded vectorization engine.
Author-email: Tu Nombre <tu_correo@ejemplo.com>
Project-URL: Homepage, https://github.com/tu_usuario/vectify
Project-URL: Bug Tracker, https://github.com/tu_usuario/vectify/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Classifier: Topic :: Multimedia :: Graphics :: Graphics Conversion
Requires-Python: >=3.8
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: opencv-python>=4.8.0
Requires-Dist: numpy>=1.24.0
Requires-Dist: onnxruntime>=1.15.0
Dynamic: license-file

# 🚀 Vectify V8 Engine

**Vectify** is an enterprise-grade, high-performance Python library that combines Artificial Intelligence (ONNX) with advanced mathematical tracing to convert raster images (JPG, PNG) into pixel-perfect, scalable vector graphics (SVG).

Built for speed and reliability, it features dynamic hardware acceleration, parallel processing, and out-of-memory (OOM) protections, making it capable of handling everything from minimal logos to highly complex 64-color photographs.

## ✨ Key Features

* **🏎️ Extreme Performance:** Utilizes OpenCV's T-API (UMat) for GPU-accelerated image refinement and CPU Multithreading for simultaneous color-layer tracing.
* **🧠 AI Super-Resolution:** Integrates an ONNX-based upscaling pipeline with Dynamic Batching to saturate GPU/CPU cores efficiently.
* **🛡️ Enterprise Stability:** Built-in chunking algorithms prevent Out-of-Memory (OOM) errors on massive images. Features automatic edge-case fallback and smart auto-resizing.
* **👻 Native Alpha Pipeline:** Flawlessly preserves PNG transparency from end to end. The background remains perfectly hollow in the final SVG.
* **🎨 Color Freedom:** Break the 16-color limit. Vectorize complex illustrations with up to 64 strict colors with zero noise.

## 📦 Installation

```bash
pip install vectify
