Deblurring Images

Deblurring Images

Matrices, Spectra, and Filtering

Per Christian Hansen, James G. Nagy, and Dianne P. O'Leary

Fundamentals of Algorithms 3, SIAM, Philadelphia, 2006

This site contains material related to the above book. In addition to the images and software used in the book, we provide additional material that can be used, e.g., for teaching a course based on the book.

Corrections

Here we provide a short list of corrections to the book (updated November 20, 2011).

HNO Functions

"HNO Functions" is a small MATLAB package, written by us, which implements all the image deblurring algorithms presented in the book. It requires MATLAB version 6.5 or newer versions. The package also includes several auxiliary functions, e.g., for creating point spread functions. The current version of the package is Versions 1.1 of April 27, 2007. To install the package, unzip the HNO.zip file and place the the directory in a convenint location. Then augment MATLAB's path to include this directory.

Challenge Files

The book includes 18 Challenges for testing the reader's understanding of the material, and the following files are needed for these Challenges. To use these files, unzip the Challenges.zip file and place the directory in a convenint location (or put the files in the same directory as the HNO Functions). Augment MATLAB's path to include this directory, if necessary.

Additional Material

Background material. We provide background material about useful matrix decompositions, many of which are used in our book. Also included is a derivation of the results in equation (7.9).

Slides. We provide slides for some of the Chapters; these slides were written jointly by P. C. Hansen, J. G. Nagy and D. P. O'Leary: Chapter 1, Chapter 3, Chapter 4, Chapter 5, Chapter 6.

External links. DIPimage and DIPlib is a combined image processing toolbox/library that could be useful to people who don't have MABLAB's image processing toolbox. The library package provides C codes for basic image display and manipulation, with a MATLAB interface. DIP image databases is a collection of links to databases of test images.

Additional Challenges

Here we provide some additional challenges related to the book. We invite readers to contribute by sending us new challenges and images, which will be included here.