VQ Compression

Image compression application with the lossy vector quantization (VQ) data compression technique.

Table of Contents

• General Info
• Graphical User Interface
• Results
• Technologies
• Setup
• Sources

General Info

Vector quantization procedure:

1. Codebook creation with prototype vectors
2. Encoding: association of the image vectors to the codebook using distance criteria
3. Decoding using look up table

Graphical User Interface

• Add Image: image that will be compressed
• Add Training: training image for better compression result (optional)
• Refinement: number of iteration for the re-finement process on the training image (max. 60)
• Delete: deletes added images from the application
• Compress: starts the compression process

Results

With Training Image

Without Training Image

Technologies

• Language: Java
  • Library: OpenCV

Setup

To run this application you need Java and to add the OpenCV JAR to your preferred IDE. You can find a tutorial to install OpenCV here.

Sources

• Java
• OpenCV