chapter 9, Image Compression

Excerpt

This chapter presents the fundamental concepts behind image compression. Images contain an enormous amount of data that must be stored and manipulated. Image compression is based upon the removal of any redundant data that may be present within an image. This chapter is divided into three sections. The first section presents the fundamentals of image compression, giving the background required to understand the next two sections. The second section discusses error-free image compression and is based upon compressing an image so that when it is uncompressed, the original image is obtained. The final section discusses lossy compression methods, which use information about the visual limitations of the human eye to remove information that will not be noticeable from an image. Uncompression of a lossy compressed image does not yield the original image but a slightly degraded version of it.

9.1 Compression Fundamentals

The amount of data required to store an image can become quite enormous for some image processing applications. For example, consider the storage of NTSC color video for a two hour movie in which the three RGB color components have been digitized. If each color component is represented by 8 bits per pixel with an image size equal to 640 × 480, a single frame of video will require 3×640×480 = 921,660 bytes of data. NTSC video produces 30 frames per second, which yields a total storage requirement for a two hour movie of 2×10¹¹ bytes of data. This is an enormous amount of data to store. Without image compression, the digital storage of NTSC color video would not be possible.