Mpeg compression is lossy and asymmetric with the

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: is encoded in the order in which it is scanned. 2. Progressive. In this mode, the image is encoded in multiple passes, so that the entire content can be visualized in a rough-to-clear process (i.e., refined during succeeding passes). This capability is convenient for browsing applications where a low-quality (low-resolution) image is more than adequate for things like scanning through the pages of a catalog, and when the desired page is reached, only that page can be subjected to subsequent passes to display it more clearly (in high-resolution). 3. Hierarchical. In this mode, the image is encoded at multiple resolutions to accommodate different types of displays. 4. Lossless. In this mode, the image is encoded in such a way that the original quality of the image can be fully restored. JPEG uses DCT encoding and quantization for the first three modes, and uses predictive encoding for the lossless mode. Since quantization is a lossy algorithm, the first three modes of encoding are lossy. The basic steps of lossy JPEG compression (shown in Figure 19.10) are as follows (the decoder performs the inverse operations): 1. If the image to be compressed is color, RGB color space is transformed to YUV color space and chrominance component (the color) is downsampled to exploit color redundancy. If the image is grayscale, this step is not needed. 2. The image is then divided into a series of blocks of 8 x 8 pixels. 3. DCT transformation is then applied to each such 8x8 block of pixels to exploit spatial redundancy. 4. The resulting DCT coefficient (the 64 frequency component in each block) are then scalar quantized using a quantization-table (O-table). 5. The resulting 2-D output of quantization is then ordered in a zigzag sequence, creating a bit stream with zero value coefficients close to the end, since the lowfrequency component (most likely nonzero) are placed in front of the highfrequency components (most likely zero). 6. Run-length encoding is then applied to the zeros of the zigzag sequence. 7. Finally, variable-length e...
View Full Document

This document was uploaded on 04/07/2014.

Ask a homework question - tutors are online