Browse Books

Reviewer: Edgar Weippl

This comprehensive work covers all aspects of multimedia databases. After the introductory chapter 1, chapter 2 deals with various multimedia data types and formats in considerable detail. Compression algorithms, which are highly relevant to multimedia, are explained thoroughly. Lu begins by explaining Huffman, run-length, and LZW coding, and then points to fractal image coding and wavelet compression. Chapter 3 briefly focuses on multimedia database design issues. Efficient retrieval is essential, especially as multimedia databases tend to store huge amounts of data. In chapters 4 through 8, various aspects of information retrieval and indexing are presented. Data structures such as B, B+, k-d, and R* trees are described in chapter 9. The rest of the book (chapters 10 through 12) highlights the relationship between operating systems and network protocols, and between performance and effectiveness issues of multimedia database systems. Existing multimedia database management systems and their possible future applications are briefly outlined in the last chapter. The language is clear and easy to understand, and encourages the reader to continue. The index could be improved, as some terms are not included. For example, the term “RGB” (for red, blue, green) is missing, although algorithms for converting between various color spaces are described in great detail. Lu identifies senior-level students, system developers, and researchers as the book's target groups. However, the book is not perfectly suited for students unless they are guided by an instructor. Some sections drift too far away from the topic of multimedia databases. Chapter 4, for example, explains basic text retrieval techniques in great detail. These basic facts are important if one is concerned with text retrieval, but the book cannot compete with books that focus on this subject, such as the excellent work by Baeza-Yates and Ribeiro-Neto [1]. Similarly, chapter 9, which deals with data structures, has little connection to the rest of the book. Apart from elaborating on unnecessary details, the author never explains why these data structures are relevant for multimedia databases or which real-world systems use them. This criticism is particularly aimed at the description of all the different RAID levels. Of course, fast and secure disk systems are important for database systems, but they are not what students want to learn about when reading a book on multimedia databases. This lack of focus makes it a challenge to use this book as a textbook, but makes it a good reference for researchers. For example, when one needs to look up how JPEG compression works, the book gives enough details to recall it. It also provides information on how deletion in R trees works, on what RAID-5 is, on various scheduling algorithms for disk access, and so on. Despite the fact that I cannot recommend the book for teaching, advanced undergraduate students can obtain a lot of general computer science knowledge by reading it. Furthermore, it is definitely nice to have as a reference, and I would not want to be without it in my daily work.