How maheshkumar-cv/bipolar-fuzzy-based-low-light-image-enhancement works
This repository provides a bipolar-fuzzy low-light image enhancement library and a Streamlit demo. In the demo, a user uploads or selects an image and chooses membership functions, fusion, an objective, and optional post-processing. The app invokes the enhancement pipeline, then presents the selected result and the negamma search history. The library also exposes command-line variants and independent entropy-based helpers. The README documents the internal stages and capabilities; the sampled excerpts show the app-to-library call, but not all internal implementation calls.
Main entry points: bfi.py Enhancement entry points · app.py Streamlit interface
Components 17
Interactive demo
- Streamlit interface
app.py - Image loading
app.py - Pipeline visualizations
app.py
Enhancement pipeline
- Enhancement entry points
bfi.py - RGB–HSI conversion
bfi.py - Membership generators
bfi.py - Negamma search
bfi.py - Image fusion
bfi.py - Intensity post-processing
bfi.py
Quality and helpers
- Scoring objectives
bfi.py - PIQE metric
bfi.py - NIQE metric
bfi.py - Entropy-based IFI
bfi.py
Image results
- Enhanced RGB image
bfi.py - Search history
bfi.py - Display and download
app.py
Other components
- Image user
Connections 18
- Image user → Image loading: uploads image
app.py - Image loading → Streamlit interface: provides image
app.py - Streamlit interface → Enhancement entry points: invokes bfi_general
app.py - Enhancement entry points → RGB–HSI conversion: converts channels
README.md - Enhancement entry points → Membership generators: selects generators
README.md - Enhancement entry points → Negamma search: searches parameters
README.md - Negamma search → Image fusion: evaluates candidates
README.md - Negamma search → Scoring objectives: scores candidates
README.md - Scoring objectives → PIQE metric: uses no-reference metric
README.md - Scoring objectives → NIQE metric: uses no-reference metric
README.md - Enhancement entry points → Intensity post-processing: optionally processes intensity
README.md - Enhancement entry points → Enhanced RGB image: returns enhanced image
README.md - Enhancement entry points → Search history: records search history
app.py - Enhancement entry points → Entropy-based IFI: also exposes IFI helpers
README.md - Streamlit interface → Pipeline visualizations: renders channels and search
app.py - Streamlit interface → Display and download: displays and downloads result
app.py - Enhanced RGB image → Streamlit interface: returns enhancement
app.py - Search history → Streamlit interface: returns candidate scores
app.py