Underwater Image Enhancement Benchmarking (FYP)

Project Overview

This repository contains the codebase for my Final Year Project, focusing on the evaluation and benchmarking of traditional underwater image enhancement techniques. The primary objective is to establish a rigorous performance baseline using physical-model and histogram-based algorithms before transitioning to deep learning and hardware optimization (Jetson Orin Nano).

Evaluated Methods:

• CLAHE (Contrast Limited Adaptive Histogram Equalization) - applied in CIELAB color space.
• UCM (Underwater Color Model) - global color correction and HSV equalization.

Evaluation Metrics:

• UCIQE (Underwater Color Image Quality Evaluation)
• UIQM (Underwater Image Quality Measure)

---

Repository Structure

Underwater_Project/
├── data/
│   ├── raw/             # (Ignored in Git) Contains UIEB dataset
│   └── processed/       # (Ignored in Git) Output for enhanced images
├── notebooks/
│   └── 01_histogram.ipynb   # CLAHE & UCM benchmarking experiments
├── utils/
│   ├── filters.py           # Core enhancement algorithms
│   └── metrics.py           # UIQM & UCIQE mathematical implementations
├── .gitignore               # Excludes heavy datasets from version control
└── README.md                # Project documentation