This project implements image segmentation and classification using a U-Net model and a custom classification model. The goal is to accurately segment images into different classes and classify them accordingly.

The project consists of the following components:

1. Data Preparation: The images and corresponding masks are loaded from the provided dataset. The images are preprocessed, including gamma correction, and converted to RGB format. The masks are one-hot encoded based on a specified colormap.

2. U-Net Model: The U-Net architecture is used for image segmentation. The model is trained on the prepared data to segment the images into different classes based on the provided masks.

3. Custom Classification Model: A custom classification model is implemented to classify the segmented images into specific classes. The model is trained on the segmented images to perform classification.

4. Evaluation: Various evaluation metrics, such as confusion matrix and accuracy score, are used to assess the performance of the segmentation and classification models.

1. Data Preparation: Run the Prepare_Data class to load and preprocess the images and masks. Specify the image and mask paths accordingly. Adjust the gamma correction and color conversion as needed.

2. Model Training: Train the U-Net model for image segmentation and the custom classification model for classification. Specify the model architectures, loss functions, optimizers, and other hyperparameters. Use the prepared data for training.

3. Model Evaluation: Evaluate the performance of the trained models using appropriate evaluation metrics. Generate and visualize the confusion matrix to assess the classification accuracy.

The project requires the following dependencies:

• Python (version 3.11.22)
• OpenCV (version 4.7.0)
• NumPy (version 1.23.5)
• Matplotlib (version 37.1)
• Scikit-learn (version 1.2.2)
• TensorFlow (version 2.12.0)
• Keras (version 2.12.0)

The project uses a provided dataset containing images and corresponding masks. The dataset should be placed in the specified directories (image_path and mask_path) before running the code.

This project is licensed under the MIT License.

• This project is inspired by the U-Net architecture for image segmentation.
• The provided dataset is sourced from https://drive.google.com/drive/folders/14DtM2x4UfIQR0au6LDARSb8YaLuagczJ?usp=sharing.