saw_mill_knot_detection/README.md

Saw Mill Knot Detection

This repository contains a complete wood defect detection system with a Tkinter-based annotation GUI and separate training/deployment scripts. Supports multiple model frameworks (RF-DETR, RT-DETR, YOLOv6, YOLOX) and is optimized for deployment on OAK-D cameras.

🎯 Project Overview

• Models: RF-DETR, RT-DETR, YOLOv6, YOLOX (all MIT/Apache 2.0 licensed)
• Dataset: 20,276 wood surface defect images
• Annotation GUI: Tkinter desktop app for manual annotation
• Training Scripts: Separate Python scripts for model training
• Deployment: OAK-D camera optimization with OpenVINO conversion
• License: All models free for commercial use

📊 Dataset Information

Source: Kaggle Wood Surface Defects Dataset

Classes (10 total):

• Live knot, Dead knot, Knot with crack, Crack, Resin
• Marrow, Quartzity, Knot missing, Blue stain, Overgrown

Dataset Split:

• Train: 16,220 images
• Valid: 2,027 images
• Test: 2,029 images

Formats Available:

• dataset_coco/ → COCO format for RF-DETR
• dataset_yolo/ → YOLO format for YOLOX, YOLOv6, YOLOv8

🚀 Quick Start

1. Environment Setup

# Clone the repository
git clone git@143.244.157.110:dillon_stuff/saw_mill_knot_detection.git
cd saw_mill_knot_detection

# Create virtual environment
python -m venv .venv
source .venv/bin/activate

# Install dependencies
pip install -r requirements.txt

2. Run the Annotation GUI

./run_tk_gui.sh --images-dir IMAGE/
# or
python tk_annotation_gui.py --images-dir IMAGE/

Auto-label requires Ultralytics for YOLO/RT-DETR weights:

pip install ultralytics

2. Setup Datasets

# Download dataset from Kaggle (requires Kaggle API)
kaggle datasets download -d kirs0816/wood-surface-defects
unzip wood-surface-defects.zip

# Create multi-format datasets
python split_coco_dataset.py  # Creates dataset_yolo/
python setup_datasets.py      # Creates dataset_coco/ and updates configs

3. Launch Annotation GUI

python tk_annotation_gui.py
# or
./run_tk_gui.sh

The Tkinter GUI supports image navigation, autosave annotations, and optional auto-label.

4. Train Models

Use the dedicated training script for all frameworks:

# Prepare dataset from annotations (optional)
python train_model.py --prepare-dataset --images-dir IMAGE --annotations annotations.json --dataset dataset_prepared

# Train models with different frameworks
python train_model.py --framework rf-detr --dataset dataset_prepared --output runs/rfdetr_training --model-size medium --epochs 50
python train_model.py --framework rtdetr --dataset dataset_prepared --output runs/rtdetr_training --model-size small --epochs 30
python train_model.py --framework yolox --dataset dataset_prepared --output runs/yolox_training --model-size nano --epochs 50
python train_model.py --framework yolov6 --dataset dataset_prepared --output runs/yolov6_training --model-size nano --epochs 50

# See TRAINING_README.md for detailed training options

5. Convert for OAK-D Deployment

# Convert trained model for edge deployment
python convert_for_deployment.py --model runs/training/weights/best.pt --output oak_d_deployment --img-size 640

# See OAK_D_WORKFLOW_README.md for complete labeling, training, and deployment workflow

📁 Project Structure

saw_mill_knot_detection/
├── tk_annotation_gui.py       # Tkinter desktop annotation GUI
├── run_tk_gui.sh              # Convenience launcher
├── train_model.py             # Unified training script for all frameworks
├── convert_for_deployment.py  # Model conversion for OAK-D deployment
├── OAK_D_WORKFLOW_README.md   # Complete workflow guide for OAK-D deployment
├── TRAINING_README.md         # Detailed training and deployment guide
├── setup_datasets.py         # Multi-format dataset setup script
├── split_coco_dataset.py     # Dataset splitting utility
├── config.py                 # Configuration settings
├── dataset_coco/             # RF-DETR dataset (COCO format)
│   ├── train/
│   │   ├── *.jpg             # Training images
│   │   └── _annotations.coco.json
│   ├── valid/
│   │   ├── *.jpg             # Validation images
│   │   └── _annotations.coco.json
│   └── test/
│       ├── *.jpg             # Test images
│       └── _annotations.coco.json
├── dataset_yolo/             # YOLOX/YOLOv6/YOLOv8 dataset (YOLO format)
│   ├── train/
│   │   ├── images/           # Training images
│   │   └── labels/           # YOLO format labels
│   ├── valid/
│   │   ├── images/           # Validation images
│   │   └── labels/           # YOLO format labels
│   ├── test/
│   │   ├── images/           # Test images
│   │   └── labels/           # YOLO format labels
│   └── data.yaml             # YOLO dataset configuration
├── runs/                     # Training outputs (excluded from git)
├── bbox_coco_dataset.json     # Original COCO annotations
├── requirements.txt           # Python dependencies
├── .gitignore                # Excludes large data files
└── README.md                 # This file

🤖 Framework Comparison

Framework	Accuracy	Speed	Memory	Deployment	Best For
RF-DETR	⭐⭐⭐⭐⭐	⭐⭐	⭐⭐⭐	CPU/GPU	Highest accuracy, research
YOLOX	⭐⭐⭐⭐	⭐⭐⭐⭐	⭐⭐⭐	Edge devices	Balanced performance
YOLOv6	⭐⭐⭐	⭐⭐⭐⭐⭐	⭐⭐	Mobile/Edge	Fast inference, production

🛠️ Usage Guide

Annotation GUI Features

The Tkinter annotation GUI provides:

• Image Navigation: Browse through dataset with current index display
• Auto-Labeling: One-click defect detection using trained YOLOX model
• Manual Annotation: Draw bounding boxes for corrections
• Real-time Visualization: Immediate display of detection results
• Export Options: Save annotations in multiple formats

Training

# Basic training
python train_yolox.py --dataset-dir dataset_split --model yolox-nano --epochs 10

# Advanced training with custom parameters
python train_yolox.py \
  --dataset-dir dataset_split \
  --model yolox-nano \
  --epochs 20 \
  --batch-size 8 \
  --img-size 640

Inference

from ultralytics import YOLO

# Load trained model
model = YOLO('runs/yolox_training/training/weights/best.pt')

# Predict on image
results = model.predict('path/to/image.jpg', conf=0.4)

# Process results
for result in results:
    boxes = result.boxes  # Bounding boxes
    for box in boxes:
        cls = int(box.cls)  # Class index
        conf = float(box.conf)  # Confidence score
        xyxy = box.xyxy.tolist()[0]  # Box coordinates

🔧 Configuration

Key settings in config.py:

DEFAULT_MODEL_WEIGHTS = "runs/yolox_training/training/weights/best.pt"
DEFAULT_IMAGES_DIR = "IMAGE/"
WOOD_DEFECT_CLASSES = [
    'Live knot', 'Dead knot', 'Knot with crack', 'Crack',
    'Resin', 'Marrow', 'Quartzity', 'Knot missing',
    'Blue stain', 'Overgrown'
]

📈 Model Performance

YOLOX-nano Results (5 epochs):

• mAP50: 0.612
• mAP50-95: 0.357
• Precision: 0.68
• Recall: 0.55

🎯 Deployment on OAK-D

The trained model can be exported for OAK-D deployment:

from ultralytics import YOLO

# Load and export model
model = YOLO('runs/yolox_training/training/weights/best.pt')
model.export(format='onnx')  # Export to ONNX for OAK-D

🤝 Contributing

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Test thoroughly
5. Submit a pull request

📄 License

This project uses the Kaggle Wood Surface Defects dataset. Please refer to the original dataset license for usage terms.

🙏 Acknowledgments

• Kaggle for providing the wood surface defects dataset
• Ultralytics for the YOLO framework