IP and CNN

A College Collaboration Project on Image Processing & CNNs

📌 Project Overview

This repository contains a collaborative effort to explore the intersection of classical Image Processing and Convolutional Neural Networks (CNNs). Our goal is to develop a pipeline that cleans and augments raw image data before feeding it into a custom deep-learning model for classification/detection tasks.

Core Objectives:

• Implement custom filters (Sobel, Gaussian, Laplacian) using NumPy/OpenCV.
• Develop a CNN architecture optimized for [mention dataset, e.g., CIFAR-10 or custom data].
• Compare performance between raw image inputs and pre-processed feature-mapped inputs.

---

🛠 Tech Stack

• Language: Python
• Deep Learning: TensorFlow / Keras (or PyTorch)
• Image Processing: OpenCV, NumPy, Matplotlib

---

🚀 Getting Started

Prerequisites

Ensure you have a C++ compiler and Python environment set up. If you are using a dedicated GPU ensure CUDA and cuDNN are correctly configured.

Installation

1. Clone the repository:

   Bash (Using the https):

   git clone https://github.com/ASAnonymousS/IP-and-CNN.git
   cd IP-and-CNN

   Bash (Using the ssh): Recommended

   git clone git@github.com:ASAnonymousS/IP-and-CNN.git
   cd IP-and-CNN

2. Create a virtual environment:

   Linux:

   python -m venv venv
   source venv/bin/activate

   Windows:

   python -m venv venv
   venv\Scripts\activate

3. Install dependencies:

   Bash

   pip install -r requirements.txt

---

📂 Project Structure

Plaintext

.
├── References/
│   ├── Image_Samples/      # Standard test images (Lenna, Pout, etc.)
│   ├── MATLAB_Files/       # DSP & Signal Processing toolkit (.m scripts)
│   └── matplotlib/         # Python scripts for color space & histogram analysis
├── README.md               # Project documentation
└── requirements.txt        # Python dependencies

🛠 Reference Toolkit Usage

1. MATLAB Setup

1. Navigate to: cd References/MATLAB_Files
2. Open scripts like binaryimg.m or IMG01.m in MATLAB to perform initial image manipulations.

2. Matplotlib & Python Analysis

1. Navigate to: cd References/matplotlib
2. Run scripts for specific analysis:
   • color_seperations.py: Visualize RGB channels.
   • histogram.py: Generate pixel intensity distributions.
   • rgb2gray_for_matplotlib.py: Custom grayscale conversion logic.

---

🧪Ongoing Development

Our team is currently focused on mastering and implementing the following frameworks to build the end-to-end pipeline:

• Data Visualization: Learning Matplotlib for EDA and results plotting.
• Numerical Computing: Mastering NumPy for high-performance matrix operations.
• Computer Vision: Implementing image filtering and transformations with OpenCV.
• Deep Learning: Developing and training models using TensorFlow, Keras, and PyTorch.

---

🤝 Contributing

Since this is a collaboration, please follow the workflow to maintain project integrity:

1. Pull the latest changes from main.
2. Create a new branch for your feature (git checkout -b your-name).
3. Environment Safety: Ensure your local virtual environment is not tracked. If you create a new environment, verify that .venv/ or venv/ is listed in the .gitignore file before staging changes.
4. Commit your updates with clear, descriptive messages (and use GPG/SSH signing if configured).
5. Push and open a Merge/Pull Request with detailed explanation of the updates in the pull request.