# Firefly Algorithm (FA) for Image Segmentation
This repository contains an implementation of the Firefly Algorithm (FA) for image segmentation using clustering. The algorithm is applied to segment an image into clusters based on pixel intensities, effectively grouping similar regions together.
## Features
- **Firefly Algorithm**: A bio-inspired optimization technique used for clustering.
- **Image Processing**: Includes grayscale conversion, histogram equalization, and image reshaping for clustering.
- **Customizable Parameters**: Easy adjustment of the number of clusters and algorithm parameters.
- **Visualization**: Outputs the segmented image and a plot of the algorithm's cost evolution.
## Parameters
You can customize the following parameters in the script:
- `k`: Number of clusters (default is 6).
- `MaxIt`: Maximum number of iterations for the Firefly Algorithm.
- `nPop`: Number of fireflies in the population.
- `gamma`: Light absorption coefficient.
- `beta0`: Attraction coefficient base value.
- `alpha`: Mutation coefficient.
## How It Works
1. **Image Preprocessing**:
- Converts the input image to grayscale.
- Adjusts contrast using histogram equalization.
- Reshapes the image into a vector for clustering.
2. **Firefly Algorithm**:
- Initializes a population of fireflies with random positions (cluster centers).
- Iteratively updates positions based on light intensity (cost function).
- Minimizes the within-cluster distance to find optimal clustering.
3. **Segmentation**:
- Maps clustered indices back to the original image shape.
- Creates a segmented image by assigning cluster labels to pixels.