REVIEW - Mathematics of Digital Images - Creation, Compression, Restoration, Recognition

Title:

Mathematics of Digital Images - Creation, Compression, Restoration, Recognition

Author:

S. G. Hoggar

ISBN:

0521780292

Publisher:

Cambridge University Press (2006)

Pages:

854pp

Reviewer:

Francis Glassborow

Reviewed:

October 2006

Rating:

★★☆☆☆

The subtitle of this book from Cambridge University Press is 'Creation, Compression, Restoration, Recognition'. As the title indicates, the book is heavy on mathematics and those who struggled to achieve success at High School math (or GCSE in the UK) will likely be drowned by the requirements of this book.

The tough mathematics requirement is unavoidable, what could have been avoided is the academic writing style. I wish that academic authors would take on a co-author whose job would be to make the contents more accessible by, for example, breaking excessively long sentences into shorter ones. This author adopts the style of writing in the first person plural. That is certainly an improvement on those who insist on writing in the third person passive.

The back cover mentions 'pseudocode', do not be led into expecting that there will be much to help you transfer theory into practical code that you can use to help your understanding. You will have to provide the understanding and probably write the pseudocode for yourself as part of the journey towards enlightenment.

There are some surprising sections in the book (at least surprising to those not already familiar with the computational needs of computing with digital images).

90 pages on matrices will surprise no one, but more than 180 pages on probability is likely to be a major surprise.

This book succeeds in its objective and if you need to understand all the intricacies of handling digital images it is about as good as you can get on the theory side. For practical programmers I would like to have seen the material supplemented with far more pseudocode. Even better would have been real compilable code in some widely available programming language. My preference would be for C++ but Python, C, C# or even Java would be acceptable.

This is a good book if you can cope with the mathematics and have the time to study the contents properly. To some extent the individual sections stand alone, but you would need to read a section linearly and not try to dip in in search of something relevant to a current programming problem. This is a book to develop your overall knowledge, understanding and skills.


Book cover image courtesy of Open Library.