How elerac/polanalyser works
Project type and purpose
• Polanalyser is an open-source Python library/tool for polarization image analysis.
File structure
• Top-level:
• LICENSE, README.md, setup.py
• documents/ (scripts to build docs/assets, notation definitions, static images)
• examples/ (client scripts demonstrating library usage)
• polanalyser/ (the package)
• polanalyser/ sub-packages and modules:
• __init__.py (public API)
• container.py (data containers)
• demosaicing.py
• stokes.py
• mueller.py
• visualization.py
• io/ (image_io.py, json_io.py, README.md)
• sympy/ (symbolic versions using SymPy)
README for insights
• Describes four main feature areas:
a. Demosaicing
b. Stokes vector analysis
c. Mueller matrix analysis
d. Visualization (numeric & symbolic)
• Lists dependencies: NumPy, OpenCV, matplotlib, SymPy (optional).
Components & Interactions
1. Client / Example Scripts
2. polanalyser API Layer (polanalyser/__init__.py)
3. IO Subsystem (polanalyser/io/image_io.py, json_io.py)
4. Core Computation Modules:
– Demosaicing
– Stokes Calculation
– Mueller Calculation
5. Visualization Module
6. Symbolic Computation Layer (polanalyser/sympy)
7. External Dependencies (NumPy, OpenCV, matplotlib, SymPy)
8. File System / Dataset Storage
• Client → API Layer (e.g., calls pa.demosaicing())
• API Layer → IO Subsystem (loads raw images/metadata)
• IO → Core Computation Modules (passes arrays/matrices)
• Core Modules ↔ External Libraries (NumPy operations, cv2)
• Visualization → File System or Display
• API Layer → Symbolic Module (if user calls symbolic API)
• Symbolic Module ↔ SymPy
c. Architectural patterns & principles:
• Layered architecture (Presentation → API → IO → Core → Visualization)
• Modular design: each optical computation in its own module
• Optional extension (SymPy) without polluting core
• Clear separation of concerns for maintainability
d. Key technologies/frameworks:
• Python 3.x
• NumPy (numerical computing)
• OpenCV (image I/O, processing)
• Matplotlib (plotting, colormaps)
• SymPy (symbolic algebra)
• setuptools / setup.py (packaging)
Detail level checklist
• Sub-packages under polanalyser: demosaicing, stokes, mueller, io, visualization, sympy.
• Interactions with external libs (NumPy, OpenCV, matplotlib, SymPy).
• Example scripts as clients invoking the API.