Wavelet transforms and image analysis

This is my implementation of one of the longer projects in Rouben Rostamian's Programming Projects in C: 1D and 2D Haar wavelet transforms applied to image compression and analysis.

Project Overview

The goal of this project is to implement Haar wavelet transforms and then apply them to image data for compression and analysis.

Features

The project is not yet complete but should have the following features:

• 1D and 2D Haar Wavelet Transforms: This project provides implementations for both 1D and 2D Haar wavelet transforms. The 1D transform can be applied to signals, while the 2D transform is designed for images.

• Image Import and Export: The project includes functions for reading and writing PGM and PPM image files. This enables you to process existing images or save the results of your wavelet analysis.

• Compression: You can apply the Haar wavelet transform to an image and discard coefficients to achieve compression while controlling the truncation error.

• Analysis: The code also allows you to analyze the frequency content of an image using the wavelet transform.

Getting Started

To get started with this project, you'll need to set up CMake and Netpbm. You can then compile the code and run e.g. the image-analysis executable to apply wavelet transforms and analyze images. Each folder corresponds to a project chapter in the book. They are self-contained and meant to be built separately. Full details are of course in the book.

Usage

Here's an example of how to use the image-analysis executable to compress a grayscale image:

$ ./image-analysis 0.9 profile.pgm profile-9.pgm

Examples

Original image

Compressed image

Acknowledgments

Rouben Rostamian for the inspiration and project idea.