Shape Recognizer

A Python library for recognizing geometric shapes from noisy measurement data, developed as part of the Algebraic Structures course (2IMA10) of the University of Technology of Eindhoven.

Overview

This project implements algorithms for identifying:

• Right-angled triangles from measured side lengths
• Regular n-gons from vertex coordinates (two methods: angular sorting and power sum invariants)

All recognizers account for bounded measurement error using mathematically derived tolerance bounds.

Project Structure

├── src/shape_recognizer/       # Main package
│   ├── geometry.py             # Point2D, ShapeData models
│   ├── main.py                 # CLI entry point
│   ├── utils.py                # Data loading utilities
│   └── recognizers/            # Recognition algorithms
│       ├── base.py             # Abstract base class
│       ├── triangle.py         # Right triangle recognizer
│       └── polygon.py          # Regular polygon recognizers
├── tests/                      # Unit tests (pytest)
├── Docs/                       # Documentation
└── report/                     # Project report

Quick Start

# Install
pip install -e .
pip install -r requirements.txt

# Run CLI
python -m src.shape_recognizer.main --help

# Run tests
pytest

Documentation

Document	Description
Docs/USAGE.md	Installation and usage guide
Docs/recognizing_shapes.md	Original project assignment
Docs/contracts.md	Coding principles and design contracts
report/report.md	Full project report with mathematical foundations

License

See LICENSE.txt.