Wavelet based Image Reconstruction from Gradient Data

A toolbox for reconstructing images from gradient data.

Wavelet based Image Reconstruction from Gradient Data
by: I.Sevcenco, P.Hampton, P.Agathoklis
Version 1.0
October 2014

This toolbox can be used to reconstruct an image (2-D signal) from gradient (first order derivative) data. Reconstructing the signal from the gradient domain is a required step in applications based on gradient manipulations, such as high dynamic range compression, image stitching, image editing, image fusion, etc.

We recommend that you start by looking at the Examples.pdf documentation. There are included three examples of how the main function (ImageRecH.m) can be used, depending on the context (e.g., grayscale image reconstruction, colour image reconstruction, grayscale image reconstruction from noisy gradient data). The code used to generate the examples is Examples.m

For details on the theory behind the toolbox, we refer the user to the following publications:

[1] I.S. Sevcenco, P.J. Hampton, P. Agathoklis, "A wavelet based method for image reconstruction from
gradient data with applications", Multidimensional Systems and Signal Processing, November 2013

[2] P.J. Hampton, P. Agathoklis, C. Bradley, "A New Wave‐Front Reconstruction Method for Adaptive
Optics Systems Using Wavelets", IEEE Journal of Selected Topics in Signal Processing, vol. 2, no. 5,
October 2008

[3] P.J. Hampton, P. Agathoklis, "Comparison of Haar Wavelet‐based and Poisson‐based Numerical Integration Techniques", Proceedings of Circuits and Systems (ISCAS), pp.1623‐1626, 2010