Skip to main content
main-content

Über dieses Buch

viii • The second new chapter, Chapter 6, discusses video compression. The chapter opens with a general description of CRT operation and basic analog and digital video concepts. It continues with a general discussion of video compression, and it concludes with a description of MPEG-1 and H.261. • Audio compression is the topic of the third new chapter, Chapter 7. The first topic in this chapter is the properties of the human audible system and how they can be exploited to achieve lossy audio compression. A discussion of a few simple audio compression methods follows, and the chapter concludes with a description of the three audio layers of MPEG-1, including the very popular mp3 format. Other new material consists of the following: • Conditional image RLE (Section 1.4.2). • Scalar quantization (Section 1.6). • The QM coder used in JPEG, JPEG 2000, and JBIG is now included in Sec­ tion 2.16. • Context-tree weighting is discussed in Section 2.19. Its extension to lossless image compression is the topic of Section 4.24. • Section 3.4 discusses a sliding buffer method called repetition times. • The troublesome issue of patents is now also included (Section 3.25). • The relatively unknown Gray codes are discussed in Section 4.2.1, in connection with image compression. • Section 4.3 discusses intuitive methods for image compression, such as subs- pling and vector quantization.

Inhaltsverzeichnis

Frontmatter

Introduction

Abstract
Giambattista delia Porta, a Renaissance scientist, was the author in 1558 of Magia Naturalis (Natural Magic), a book in which he discusses many subjects, including demonology, magnetism, and the camera obscura. The book mentions an imaginary device that has since become known as the “sympathetic telegraph.” This device was to have consisted of two circular boxes, similar to compasses, each with a magnetic needle. Each box was to be labeled with the 26 letters, instead of the usual directions, and the main point was that the two needles were supposed to be magnetized by the same lodestone. Porta assumed that this would somehow coordinate the needles such that when a letter was dialed in one box, the needle in the other box would swing to point to the same letter.
David Salomon

1. Basic Techniques

Abstract
Data compression is achieved by reducing redundancy, but this also makes the data less reliable, more prone to errors. Making data more reliable, on the other hand, is done by adding check bits and parity bits, a process that increases the size of the codes, thereby increasing redundancy. Data compression and data reliability are thus opposites, and it is interesting to note that the latter is a relatively recent field, whereas the former existed even before the advent of computers. The sympathetic telegraph, discussed in the Preface, the Braille code of 1820 (Section 1.1.1), and the Morse code of 1838 (Table 2.1) use simple forms of compression. It therefore seems that reducing redundancy comes naturally to anyone who works on codes, but increasing it is something that “goes against the grain” in humans. This section discusses simple, intuitive compression methods that have been used in the past. Today these methods are mostly of historical interest, since they are generally inefficient and cannot compete with the modern compression methods developed during the last 15–20 years.
David Salomon

2. Statistical Methods

Abstract
The methods discussed so far have one common feature, they assign fixed-size codes to the symbols (characters or pixels) they operate on. In contrast, statistical methods use variable-size codes, with the shorter codes assigned to symbols or groups of symbols that appear more often in the data (have a higher probability of occurrence). Designers and implementors of variable-size codes have to deal with the two problems of (1) assigning codes that can be decoded unambiguously and (2) assigning codes with the minimum average size.
David Salomon

3. Dictionary Methods

Abstract
Statistical compression methods use a statistical model of the data, and the quality of compression they achieve depends on how good that model is. Dictionary-based compression methods do not use a statistical model, nor do they use variable-size codes. Instead they select strings of symbols and encode each string as a token using a dictionary. The dictionary holds strings of symbols and it may be static or dynamic (adaptive). The former is permanent, sometimes allowing the addition of strings but no deletions, whereas the latter holds strings previously found in the input stream, allowing for additions and deletions of strings as new input is being read.
David Salomon

4. Image Compression

Abstract
The first part of this chapter discusses digital images and general approaches to image compression. This is followed by a description of about 30 different compression methods. The author would like to start with the following observations: 1. Why were these particular methods included in the book, while others were ignored? The simple answer is this: Because of the documentation available to the author. Image compression methods that are well documented were included. Methods that are kept secret, or whose documentation was not clear to the author, were left out.
David Salomon

5. Wavelet Methods

Abstract
The concept of a transform is familiar to mathematicians. It is a standard mathematical tool used to solve problems in many areas. The idea is to change a mathematical quantity (a number, a vector, a function, or anything else) to another form, where it may look unfamiliar but may exhibit useful features. The transformed quantity is used to solve a problem or to perform a calculation, and the result is then transformed back to the original form.
David Salomon

6. Video Compression

Abstract
Sound recording and the movie camera were among the greatest inventions of Thomas Edison. They were later united when “talkies” were developed, and they are still used together in video recordings. This unification is one reason for the popularity of movies and video. With the rapid advances in computers in the 1980s and 1990s came multimedia applications, where pictures and sound are combined in the same file. Since such files tend to be large, compressing them became a natural application.
David Salomon

7. Audio Compression

Abstract
Text does not occupy much space in the computer. An average book, consisting of a million characters, can be stored uncompressed in about 1 Mbyte, since each character of text occupies one byte (the Colophon at the end of the book illustrates this with precise data from the book itself).
David Salomon

8. Other Methods

Abstract
Previous chapters discuss the main classes of compression methods: RLE, statistical methods, and dictionary-based methods. There are data compression methods that are not easy to classify and do not clearly belong in any of the classes discussed so far. A few such methods are described here.
David Salomon

Backmatter

Weitere Informationen