JPEG Coding

CHEN GuowangFANG Wei

HUANG Baihan

General Coding

Source Data Source Coding Channel Coding Transmission

Reducing Redundancy

Adding Redundancy to

Ensure Correctness

JPEG Compression

Compressing each 8x8 block

Source Image Color Space Transformation Down Sample Break into 8x8

Blocks

2-D DCT Quantization Lossless Coding Compressed Image

Color Space Transformation• Y: brightness• Cb & Cr: blue-difference and red-difference

Original Down-sample all components

Preserve Y component

Compressing each 8x8 block

JPEG Compression

Source ImageColor Space

Transformation:RGB to YCbCr

Down Sample Cb, Cr

Components

Break into 8x8 Blocks

2-D DCT Quantization Lossless Coding Compressed Image

2-D Basis Images

2-D DCT

Input image block After 2-D DCT

139 144 149 153 155 155 155 155144 151 153 156 159 156 156 156150 155 160 163 158 156 156 156159 161 162 160 160 159 159 159159 160 161 162 162 155 155 155161 161 161 161 160 157 157 1571

( ,

2

)

62 16

f x y

161 163 162 157 157 157162 162 161 161 163 158 158 158

1260 1 12 5 2 2 3 123 17 6 3 3 0 0 111 9 2 2 0 1 1 07 2 0 1 1 0 0 0

( , )1 1 1 2 0 1 1 1

2 0 2 0 1 1 1 11 0 0 1 0 2 1 13 2 4 2 2 1 1 0

F u v

Quantization

÷ =

Coefficient Matrix ÷ Quantization Matrix = Result

Zigzag Scan & Run-length Coding

Numbers of Zeros Before Value

1 -2

0 -1

0 -1

0 -1

2 -1

End of Block

Q & A