2. Content :
• Concept of Image Compression.
• Image Compression Models.
• Types of Image Compression.
• Variable-length Coding.
3. Image Compression
• Refers to reducing the amount of data required to represent a digital
image.
• Image compression address the problem of reducing the amount of
data required to represent a digital image with no significant loss of
information.
4. Why…….?
• Principal objective: To minimize the number of bits required to
represent an image.
• Reducing the image storage.
• Transmission requirements.
5. Image Compression
• Image = Information + Redundant Data
• Three Principal type of Data Redundancies
used in Image Compression :
Coding
Redundancy
Interpixel
Redundancy
Psychovisual
Redundancy
6. Image Compression
The number of bits used to represent each pixel is based
on number of gray levels used to represent the image
We represent the entire image by using least possible
number of bits. In this way we can reduce the coding
redundancy.
Coding Redundancy
7. Image Compression
It is also called Spatial & Temporal redundancy.
In an image each pixel depends on its neighbors.
If spatial resolutions is high then inter pixel redundancy is
high.
Interpixel Redundancy
8. Image Compression
Certain information has relatively less importance for the
quality of image perception. This information is said to be
psychovisually redundant.
Removing this type of redundancy is a lossy process and the
lost information cannot be recovered.
The method used to remove this type of redundancy is called
quantization which means the mapping of a broad range of
input values to a limited number of output values.
Psychovisual Redundancy
9. Image Compression Model
• The image compression system is composed of 2 distinct functional
component: an encoder & a decoder.
Source
Encoder
Channel
Encoder
Channel
Channel
Decoder
Source
Decoder
Encoder Decoder
Compression
(No redundancies)
Noise tolerant representation
(additional bits are included to
guarantee detection &
correction of error due to
transmission over channel.-
Hamming Code)
10. Image Compression Model
• Encoder performs Compression while Decoder performs Decompression.
Encoder is used to remove the redundancies through a series of 3
independent operations.
Mapper Quantizer
Symbol
Encoder
Channel
No Interpixel
redundancies
(Reversible)
No
Psychovisual
redundancies
(non-
reversible)
No Coding
redundancies
(Reversible)
Encoder
11. Image Compression Model
• Inverse steps are performed .
Channel
Symbol
Encoder
De-quantizer Inverse Mapper
Decoder
12. Types of Image Compression
• Image data compression methods fall into two common categories:
Lossy
compression
Lossless
compression
13. Lossy Compression
A lossy compression method is one where compressing
data and then decompressing it retrieves data that may
well be different from the original, but is close enough to
be useful in some way.
14. Lossy Compression
Used to compress multimedia
data (audio, video, still images),
especially in applications such
as streaming media and
internet telephony.
Provide higher levels of data
reduction
Result in a less than perfect
reproduction of the original
image
15. Lossless Compression
• Also called Information preserving compression.
• Compress and decompress images without losing information.
16.
17. Variable-length Coding
• The coding redundancy can be minimized by using a variable-
length coding method where the shortest codes are assigned to
most probable gray levels.
• The most popular variable-length coding method is the Huffman
Coding.
18. Huffman Coding
• The Huffman coding involves the following steps.
1) Find the gray – level probabilities for the image by finding the
histogram.
2) Order the input probabilities (histogram magnitudes) from smallest to
largest.
3) Combine the smallest two. (add the two smallest)
4) GOTO step 2, until only two probabilities are left.
19. • Ex.
• Find 010100111100 using
Huffman.
• Find the avg no of bits
required to represent
each pixel(Lavg).
Huffman
Hinweis der Redaktion
Mapper: transforms input data in a way that facilitates reduction of inter pixel redundancies.
Quantizer: achieved by compressing a range of values to a single quantum value. When the number of discrete symbols in a given stream is reduced, the stream becomes more compressible.
Symbol encoder: assigns the shortest code to the most frequently occurring output values
Lavg = Σ l(rk) pr(rk) احتمالية كل بيت * عدد البايناري بيت لهذه الاحتمالية bits / pixel
Total no. of bits required to represent entire image = MNLavg = 256*256*L