Deblurring Images: Matrices, Spectra and Filtering by Per Christian Hansen, James G. Nagy, Dianne P. O'Leary
Paul K. Marriott · International Statistical Review · 2007
Deblurring Images: Matrices, Spectra and Filtering Per Christian Hansen, James G. Nagy, Dianne P. O'Leary SIAM , 2006 , xiv + 130 pages, US$ 63, softcover ISBN : 978-0-89871-618-4 Readership: Suitable for anyone who just wants to dabble in image analysis. This book takes a practical and computational approach to one of the basic problems of image analysis: removing the blur from an image. Using MATLAB code and plenty of realistic examples, all of which can be downloaded from the book's website, the volume takes the reader through the fundamental theory and practice of deblurring. The volume assumes a basic familiarity with linear algebra, for example the singular value decomposition, and it certainly helps if the reader has used MATLAB before. Overall, the level is suitable for mathematics or engineering undergraduate students as a course text. However the practicality and clarity of the writing means that it is also very suitable for anyone who just wants to dabble in image analysis. The text is designed around a series of ‘challenges'’, essentially short projects based on the book's code and examples. I found that these challenges allowed a non-specialist to quickly pick up the basic ideas and to master some of the practicalities. The earlier chapters start with the basics of the linear blurring model and the computational issues associated with noise and inverting poorly conditioned linear transformations. Later chapters look at different filtering and regularization techniques. Overall this book is an excellent and clearly written introductory text.