Weld Symbol Identifier

This repository contains a YOLO-based model for identifying weld symbols in engineering drawings. The model is trained to detect 16 distinct weld types commonly found in fabrication and drafting workflows. These weld types were taken from the document: Weld Symbols in Engineering Drawings

This model is intended to be used in conjunction with a weld symbol annotation tool, such as Label Studio. This is a base model that can be used to identify weld symbols in images, but it is not optimized for all use cases. The model was built using Ultralytics and YOLOv8.

Supported Weld Types

The model detects the following weld symbols (code references in camelCase):

• Backing Spacer (backingSpacer)
• Bevel Groove Weld (bevelGroove)
• Callout Tail (calloutTail)
• Callout Tip (calloutTip)
• Convex/Concave Modifier (convexConcave)
• Field Weld Symbol (fieldWeld)
• Fillet Weld (filletWeld)
• Flare Bevel Groove Weld (flareBevelGroove)
• Flush/Flat Modifier (flushFlat)
• J-Groove Weld (jGroove)
• Melt-Through Symbol (meltThrough)
• Plug Weld (plugWeld)
• Square Groove Weld (squareGroove)
• V-Groove Weld (vGroove)
• Weld-All-Around Symbol (weldAllAround)

---

Full Setup Guide (No Prerequisites Assumed)

1. Install Python

Ensure Python 3.10+ is installed. Check version:

python --version

If not installed, download from Python.org.

---

2. Install Git

Make sure Git is available to clone the repository:

git --version

If not, install from Git SCM.

---

3. Create a Virtual Environment

Recommended for dependency isolation:

python -m venv weldCalloutEnv

Activate it:

• Windows (PowerShell):

  weldCalloutEnv\Scripts\Activate

• macOS/Linux:

  source weldCalloutEnv/bin/activate

---

4. Clone the Repository

git clone https://github.com/RylanBosquez/weldCalloutIdentifier.git
cd weldCalloutIdentifier

---

5. Install Dependencies

pip install --upgrade pip
pip install -r requirements.txt

---

Usage

Run inference on example images:

python main.py

The script will:

• Load each image from ./assets/weldExamples
• Run YOLOv8 inference
• Apply non-maximum suppression to filter overlapping boxes
• Display the image with bounding boxes and weld type labels

---

Example Images

Tip, Tail, and Fillet Weld

Tip, Tail, and Bevel Groove Weld

Tip, Tail, and Fillet Weld Both Sides

Tip, Tail, and V-Groove Weld

Tip, Tail, Fillet, and Weld-All-Around

Add your test images to the ./assets/weldExamples folder.

---

Applications

• Automated annotation in CAD workflows
• Symbol classification for engineering documentation
• Integration with Label Studio or other annotation tools