Automated Graph Digitization and Performance Analysis

A Python-based tool for extracting numerical data from performance graphs (pumps & compressors), fitting curves and generating performance reports._

---

Project overview

This repository contains the implementation of an interactive application that:

• Loads graph images (PNG, JPG, BMP),
• Lets users manually trace curves (click points) on the displayed graph,
• Transforms pixel coordinates to real-world axis values based on user-supplied axis limits,
• Performs curve fitting (polynomial or spline) and interpolation,
• Calculates performance metrics (e.g., flow, head, efficiency),
• Exports results (raw & transformed points, fitted curve data, plots) to an Excel report.

Use cases: pump/compressor performance analysis, digitizing plot data from papers, lab reports and legacy documents.

---

Quick demo

# create environment
python -m venv venv
source venv/bin/activate   # On Windows: venv\Scripts\activate

# install dependencies
pip install -r requirements.txt

# run GUI app
python src/app.py

---

Repo structure

/
├─ src/
│  └─ curvetracing.py          # Main Python script containing GUI, tracing, and analysis logic
│
├─ examples/
│  ├─ demo_graph.png           # Sample input graph image
│  ├─ output_trace_1.xlsx      # Example Excel output file
│  ├─ output_trace_2.xlsx      # Another example Excel output
│  ├─ performance_data.xlsx    # Example performance analysis data
│
│
├─ requirements.txt            # Dependencies list
├─ README.md                   # Project documentation

---

Features

• Image zooming & panning for precise point selection.
• Real-time visual feedback for selected points.
• Undo / redo last point.
• Two fitting modes: polynomial (configurable degree) and cubic spline.
• Export: .xlsx containing raw pixel points, transformed values, fitted curve sampled points, and plots embedded in the sheet.

---

Installation (detailed)

1. Clone the repo:

git clone https://github.com/<your-username>/automated-graph-digitization.git
cd automated-graph-digitization

2. Create & activate a virtual environment:

python3 -m venv venv
source venv/bin/activate      # Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

---

Usage (GUI)

1. python src/app.py
2. In the GUI:
   • Load Image → select a graph image.
   • Set Axes → input X_min, X_max, Y_min, Y_max (real-world axis values).
   • Start Tracing → click points along the curve. Use Undo to remove last point.
   • Fit Curve → choose polynomial degree or spline.
   • Export Report → save .xlsx containing the results and plots.

---

Usage (CLI / script example)

You can also run non-GUI pipeline to transform a pre-collected CSV of pixel points and produce a report:

python src/transform_and_report.py --points examples/demo_points.csv \
  --image examples/demo_graph.png \
  --x-min 0 --x-max 100 --y-min 0 --y-max 10 \
  --fit polynomial --degree 3 \
  --output demo_out.xlsx

---

Tests

Run unit tests with pytest:

pytest -q

---

Contributing

• Fork the repository.
• Create a feature branch: git checkout -b feature/my-feature
• Commit your changes: git commit -m "Add new feature"
• Push: git push origin feature/my-feature
• Create a Pull Request detailing the changes and test steps.

---

TODO / Future work

• Automatic curve detection using CNN / image segmentation (ChartOCR-like models).
• Support for logarithmic and polar axes.
• Web UI + cloud processing for batch jobs.
• Digital twin integration for real-time monitoring.