Some standard multimedia compression techniques over

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: aracters appearing with the frequency shown. Note that the Huffman encoding scheme pays off only with uneven character distribution. For example, if all characters in Figure 19.9 were used equally often, the mean number of bits per character would be 3.33. This is worse than a fixedlength coding scheme in which the mean number of bits per character would be 3. For probabilities given in Figure 19.9, the average number of bits per character is 2.05. Character Probability of Code Occurrence A 0.40 0 B 0.29 10 C 0.20 110 D 0.08 1110 E 0.02 11111 F 0.01 11110 Figure 19.9. An example showing a Huffman code for a given character set with the probability of occurrence of the characters. Lossy Techniques The following two are the commonly used lossy compression techniques: 1. Predictive encoding. In many types of multimedia objects (like digital audio or video) adjacent samples or frames are generally similar to each other. The predictive encoding method takes advantage of this fact and stores only the initial sample and the difference values between every two adjacent samples for all the samples in the compressed form. The compressed data is decompressed by reproducing a sample from its previous sample and the difference value between the two samples. Since the size of the difference value between samples is usually much smaller than the size of: sample itself, the file size of a compressed data is generally much smaller than the uncompressed data file. Note that the sample may be a pixel, line, audio sample, or video frame. 2. Transform encoding. In this method, data is converted from one domain to another that is more convenient for compression. For example, data may be converted from the time domain to the frequency domain. DCT (Discrete Cosine Transform) encoding is the best example of this method. It is described below. DCT encoding transforms samples from the time domain to the frequency domin. The input data is fed as a two-dimensional block, typically with 8x8 pixels. DCT transforms an 8x8 block of pixel color values into an 8x8 matrix of spatial frequencies. The low frequencies (i.e., general information)...
View Full Document

This document was uploaded on 04/07/2014.

Ask a homework question - tutors are online