Image Compression?

The problem of reducing the amount of data required to represent a digital image. From a mathematical viewpoint: transforming a 2D pixel array into a statistically uncorrelated data set.

Why do We Need Compression?

For data STORAGE and data TRANSMISSION DVD Remote Sensing Video conference FAX Control of remotely piloted vehicle The bit rate of uncompressed digital cinema data exceeds 1 Gbps

Information vs Data
DATA = INFORMATION + REDUNDANT DATA

Why Can We Compress?

Spatial redundancy Neighboring pixels are not independent but correlated

REDUNDANT DATA

INFORMATION

Temporal redundancy

Image Compression Model.

Image Compression Model.

Reduce interpixel redundancy (Reversible) Reduce coding redundancy (Reversible)

Remove Input Redundancies

Increase the Noise Immunity

The source encoder and decoder

CLASSIFICATION
Lossless compression
lossless compression for legal and medical documents, computer programs exploit only code and inter-pixel redundancy

Error-Free Compression
Applications:
Archive of medical or business documents Satellite imaging Digital radiography

Lossy compression
digital image and video where some errors or loss can be tolerated exploit both code and inter-pixel redundancy and sychovisual perception properties

They provide: Compression ratio of 2 to 10.

Bit-plane coding

Bit-plane coding
Bit-plane coding has been widely used in lossless compression of gray-scale images or color palette images. Progressive transmission can be used in bit-plane coding strategy. P r o c e s s t h e i ma g e s b i t p l a n e s i n d i v i d u a l l y. D e c o mp o s e t h e i ma g e i n t o a s e r i e s o f b i n a r y i ma g e s . C o mp r e s s e a c h b i n a r y i ma g e v i a a b i n a r y c o mp r e s s i o n me t h o d .

Error-Free Compression
Bit-plane coding
abcdef

Bit-plane coding
The intensities of an m-bit monochrome image can be represented in the form of base-2 polynomial: am-12m-1+am-22m-2++a121+a020 Therefore , to decompose an image into a set of binary images , we need to separate m coefficient of the polynomial into m 1-bit plane. The lowest order bit-plane (corresponds to the least significant bit) is generated by a0 bits of each pixel, while the highest order bit-plane contains am-1 bits.

Bit-plane coding is based on decomposing a multilevel image into a series of binary images and compressing each binary image .

a b c d e
f

Bit-plane coding
However, this approach leads to the situation, when small changes of intensity can have significant impact on bit-planes. For instance, if a pixel intensity 127 (01111111) is adjacent to a pixel intensity 128(10000000), every bit will contain a corresponding 0 to 1 (or 1 to 0) transition. Alternatively, an image can be represented first by an m-bit Gray code. This code gm-1g2 g1 g0 corresponding to the polynomial is computed as + gi=ai ai+1 0 i m-2 gm-1=am-1 The property of this code is that the successive code words differ in only one bit position and small changes in intensity are less likely to affect all m bit-planes.

BitBit -Plane Coding

Original image Bit 7 Binary image compression Binary image compression

Bit 6

Bit 0

Binary image compression

Bit plane images

Example of binary image compression: Run length coding

Bit Planes
Bit 7 Bit 3

Gray-coded Bit Planes

Original bit planes
Gray code: a7 g7

gi = ai ai +1 for
Bit 6 Bit 2 and a6 g6

0i6

g 7 = a7
Original gray scale image ai= Original bit planes Bit 5 Bit 1 a5 g5

= XOR

Bit 4

Bit 0

a4

g4

GrayGray -coded Bit Planes (cont.)

a3 g3

Relative Address Coding (RAC)

Concept: Tracking binary transitions that begin and end back black and white run

There are less 0-1 and 1-0 transitions in grayed code bit planes. Hence gray coded bit planes are more efficient for coding.

a2

g2

a1

g1

a0

g0

Contour tracing and Coding

Represent each contour by a set of boundary points and directional's .