<<<<<<< HEAD

Pavement Distress Detection Pipeline

Overview

This project implements a YOLO-based pipeline for multi-class pavement distress detection using a combination of public datasets and self-collected video data.

Directory Structure

• Datasets/ → Public datasets and processed splits
• env_yolo/ → Python environment for YOLO training
• models/ → Model checkpoints
• scripts/ → Data processing and training scripts
• video_to_img/ → Video frame extraction utilities
• auto_labels/ → Auto-labeled outputs
• runs/ → Training logs and metrics

Workflow Summary

1. Public datasets curated and harmonized
2. Videos converted to frames
3. Initial model trained on public data
4. Auto-labeling cycles applied
5. Final model trained on merged dataset =======

Pavement_Distress

b101f0405139a3027d646ff13ebfae4635dedbb9