JPEG Compression Project

A C++ implementation of the JPEG compression and decompression pipeline using OpenCV. This project demonstrates the fundamental steps of the JPEG standard, from color space transformation to entropy coding.

Features

• Full Compression Pipeline: Implements the standard JPEG workflow (DCT, Quantization, Zig-Zag, RLE).
• Custom DCT Implementation: Manual computation of Forward and Inverse Discrete Cosine Transforms on 8x8 blocks.
• Entropy Coding: Includes Zig-Zag scanning and Run-Length Encoding (RLE) to a text-based format.
• Performance Metrics: Calculates compression rate and size reduction percentage.
• Visual Feedback: Displays original, compressed (quantized domain), and reconstructed images.

Implementation Details

The project follows these sequential steps:

1. Color Space Conversion: Transforms the image from BGR to YCrCb.
2. Padding: Ensures image dimensions are multiples of 8 for block processing.
3. Level Shifting: Subtracts 128 from pixel values to center them around zero.
4. FDCT (Forward DCT): Converts 8x8 spatial blocks into the frequency domain.
5. Quantization: Applies a standard Luminance Quantization Matrix (Q50) to discard high-frequency information.
6. Serialization:
   • Zig-Zag Ordering: Reorders 2D coefficients into a 1D sequence.
   • RLE (Run-Length Encoding): Compresses the sequence by grouping identical consecutive values.
7. Decompression: Reverses the above steps (RLE Decoding -> Dequantization -> IDCT -> Level Shifting -> BGR Conversion).

Project Structure

• OpenCVApplication.cpp: Core logic and UI menu.
• common.h/cpp: Utility functions for file handling and image resizing.
• Images/: Sample images for testing.
• rle.txt: Intermediate file storing the encoded image data.

Getting Started

Prerequisites

• Visual Studio (Solution file provided)
• OpenCV 4.x

Running the Application

1. Open OpenCVApplication.sln in Visual Studio.
2. Build the project in x64 mode.
3. Run the application.
4. Select Option 1 from the menu to test the compression on an image of your choice.

Sample Output

Upon processing an image, the console will display:

• Compression Rate: Ratio between original and compressed non-zero coefficients.
• Size Reduction: Percentage of data reduced.