scientific article; zbMATH DE number 1356545
The book under review contains the text of the doctoral dissertation of the author. As such, it is not a traditional textbook, but it gives the results of rather deep research about a specific problem in image analysis. As the title suggests, it is the aim to solve the problem of the interpretation of digital images by using Markov random fields, by working at two levels. At the low level (this name corresponds more or less to the name ``middle level, that often is used in image analysis), pixels are labeled according to their affiliation to classes of textures, and they are grouped to form regions; textures at the low level are represented by the Gibbs distribution of the Markov random field for the gray values. At the high level, again using Markov random fields, regions are labeled to obtain the meaning (interpretation, description) of the objects. The book consists of five chapters, followed by some conclusions and a list of references, for a total of less than 90 pages. Chapters 1 and 2, entitled respectively ``Introduction and ``Theoretical fundamentals, contain a general description of the problem at hand and of the way that will be used to solve it, and some mathematical background about graphs and random fields in order to understand the mathematical description of the method. Chapter 3 (Recognition of texture) and chapter 4 (Interpretation of images) form the core of the book, as they describe the two-level way and the interaction between the two levels that has been used by the author. Finally, in chapter 5 the author applies his method on different images, including multispectral serial photographs. Research of this type is a continuation of already existing methods. A short overview of such methods that have been introduced especially during the last decade, together with a description of the similarities and the differences of the methods of the author with the former ones, has been given in the introductory chapter 1. As the real practical value of such research is given by the results it affords, an investigation on test images has been given in chapter 5. From them the author formulates conclusions and suggestions for further research.
- Image textures and Gibbs random fields
- Bayesian approach to image interpretation
- Multi-label Markov random fields as an efficient and effective tool for image segmentation, total variations and regularization
- Two level image interpretation based on Markov random fields
- Signal and Image Segmentation Using Pairwise Markov Chains
This page was built for publication:
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4700464)