Home

YOLO image analysis wrapper for detecting elements and layouts in architectural design files.

• Target Framework: netstandard2.0

📚 API Reference & Guides

• DiGi.YOLO Overview
• Year Built Prediction Pipeline — Automated historical aerial orthophoto inference and detector-only dating
• Running Scripts Manually — Step-by-step guide for predict.py and train.py
• Classes — Core models (YOLOModel, Image, Label, BoundingBox, BoundingBoxResult)
• Interfaces — Core interfaces (IBoundingBox)
• Enums — Data split categories (Category.Train, Category.Validate, Category.Test)
• Constants — Directory names and file extensions

🚀 Key Features & Usage

• Year Built Prediction Pipeline: Integrates frozen YOLO detection weights into an end-to-end automated pipeline dating building construction years from historical aerial imagery.
• Runner Script Export (Modify.WriteScripts): Extract embedded Python runner scripts (train.py, predict.py, utils.py, requirements.txt) directly into any directory.
• Inference (predict.py): Configurable command-line tool supporting --model, --source, --conf, and --output options.
• Dataset Management: Automatic dataset structure layout, configuration generation (conf.yaml), image and label serialization.

🔗 Dependencies

• DiGi.Core

---

🌐 DiGi Ecosystem

• Foundational: DiGi.Core | DiGi.Math | DiGi.Unit | DiGi.Log
• Geometry & Graphics: DiGi.Geometry | DiGi.GLTF | DiGi.Rhino
• GIS & Data: DiGi.GIS | DiGi.OSM | DiGi.BDOT10k
• Simulation: DiGi.Analytical | DiGi.Solar | DiGi.Tas

Part of the DiGi software suite for BIM and CAD integrations.