JPEG-Encoding

This repository contains my implementation of a JPEG-like image compression engine in Python for learning demonstrations. The project covers the full compression and decompression pipeline, including:

• 2D Discrete Cosine Transform (DCT) & Quantization – Frequency-domain transformation to reduce redundancy.
• Huffman Encoding – Entropy compression for efficient data storage.
• Image Reconstruction – Using inverse quantization and IDCT to restore images.
• Support for Grayscale & Color Images – Converts RGB images to the YUV color space for better compression.
• Configurable Quality Factor – Adjusts compression levels to balance file size and image quality.

Installation & Dependencies

Ensure you have Python installed and install the required dependencies using:

pip install -r requirements.txt

Usage

Compress a Single Image

To compress an image using main.py, run:

python3 main.py <image_path> <grayscale> <quality_factor>

• <image_path>: Path to the image file (e.g., Pictures/building.jpeg).
• <grayscale>: Boolean (true or false) to indicate if the image should be converted to grayscale.
• <quality_factor>: Integer (1-100) to control the compression level (higher means better quality, larger file size).

Example:

python3 main.py Pictures/building.jpeg true 50

This compresses building.jpeg as a grayscale image with a quality factor of 50.

---

Evaluate Compression Quality

To visualize compression results across all quality factors (1-100) and save data to a CSV file, use results.py:

python3 results.py <image_path> <grayscale> <output_directory> <csv_output>

• <image_path>: Path to the image file.
• <grayscale>: Boolean (true or false) to convert the image to grayscale.
• <output_directory>: Folder where the results will be stored.
• <csv_output>: Path to save the CSV file containing size comparisons.

Example:

python3 results.py Pictures/building.jpeg true ./results ./results/image_quality_data.csv

This will generate visualizations and save a CSV file containing data such as:

Quality Factor	Encoded Size (KB)	Decoded Size (KB)	Original Size (KB)
10	X KB	Y KB	Z KB
50	X KB	Y KB	Z KB
100	X KB	Y KB	Z KB

---

Project Structure

│── main.py             # Main script for JPEG compression
│── results.py          # Generates visualizations and CSV reports
│── huffman.py          # Huffman encoding and decoding logic
│── dct.py              # Implements DCT and inverse DCT
│── quantization.py     # Quantization functions
│── utils.py            # Helper functions
│── requirements.txt    # Dependencies
│── README.md           # This documentation
│── Pictures/           # Folder for input images
│── results/            # Folder for output images & CSV results

Features & Future Improvements

✔️ Current Features:

• Full JPEG compression pipeline.
• Adjustable quality factor.
• Grayscale & color image support.
• Huffman-based lossless encoding.
• CSV-based storage of size comparisons.

🚀 Planned Improvements:

• None, feel free to adjust the code yourself

---

License

This project is open-source and available under the MIT License.