COMPRESSION
COMPRESSION
MODELS
MODELS
RANJANA
RANJANA
22010103043
22010103043
WHAT IS COMPRESSION ?
WHAT IS COMPRESSION ?
• Image compression is the process of reducing amount of data
Image compression is the process of reducing amount of data
required to represent or store an image.
required to represent or store an image.
• Process of encoding data so that it take less storage space or
Process of encoding data so that it take less storage space or
less transmission time.
less transmission time.
Why Do We Need Image Compression?
Why Do We Need Image Compression?
• Consider a black and white image that has a resolution of
Consider a black and white image that has a resolution of
1000*1000
1000*1000
• each pixel uses 8 bits to represent the intensity.
each pixel uses 8 bits to represent the intensity.
• So total no of bits required = 1000*1000*8 = 80,00,000 bits per
So total no of bits required = 1000*1000*8 = 80,00,000 bits per
image.
image.
• And consider if it is a video with 30 frames per second of the
And consider if it is a video with 30 frames per second of the
above-mentioned type images
above-mentioned type images
• total bits for a video of 3 secs is: 3*(30*(8, 000, 000))
total bits for a video of 3 secs is: 3*(30*(8, 000, 000))
• =720, 000, 000 bits
=720, 000, 000 bits
• Storage Efficiency: Compressed images require less storage
Storage Efficiency: Compressed images require less storage
space.​
space.​
• Transmission Speed: Smaller image files can be transmitted
Transmission Speed: Smaller image files can be transmitted
faster over networks.​
faster over networks.​
• Cost Reduction: Less storage and bandwidth usage lead to cost
Cost Reduction: Less storage and bandwidth usage lead to cost
savings.​
savings.​
• Improved Performance: Enhances performance of applications
Improved Performance: Enhances performance of applications
by reducing load times.
by reducing load times.
Data vs. Information
Data vs. Information
• Data: Raw pixel values in an
image.​
• Information: Meaningful content
derived from data.​
• Compression focuses on reducing
data redundancy while
preserving essential information.
Redundancy and Its Types
Redundancy and Its Types
•Redundancy means
Redundancy means
repetitive data or
repetitive data or
unwanwanted data
unwanwanted data
CLASSIFICATION
CLASSIFICATION
 Interpixel Redundancy
Interpixel Redundancy
Psychovisual Redundancy
Psychovisual Redundancy
 Coding Redundancy
Coding Redundancy
Coding Redundancy
Coding Redundancy
 Coding redundancy is caused due to poor selection of coding
Coding redundancy is caused due to poor selection of coding
technique
technique
 Coding techniques assigns a unique code for all symbols of message
Coding techniques assigns a unique code for all symbols of message
• Wrong choice of coding technique creates unnecessary additional
Wrong choice of coding technique creates unnecessary additional
bits. These extra bits are called redundancy
bits. These extra bits are called redundancy
• CODING REDUNDANCY = AVERAGE BITS USED TO CODE - ENTROPHY
CODING REDUNDANCY = AVERAGE BITS USED TO CODE - ENTROPHY
Interpixel Redundancy
Interpixel Redundancy
• This type of redundancy is related with the inter-pixel
This type of redundancy is related with the inter-pixel
correlations within an image.
correlations within an image.
• The value of any given pixel can be predicted from the value of
The value of any given pixel can be predicted from the value of
its neighbours or adjacent pixels that are highly corelated.
its neighbours or adjacent pixels that are highly corelated.
• Inter-pixel dependency is solved by algorithms like:
Inter-pixel dependency is solved by algorithms like:
• Predictive Coding, Bit Plane Algorithm, Run Length
Predictive Coding, Bit Plane Algorithm, Run Length
Psychovisual Redundancy
Psychovisual Redundancy
• The eye and the brain do not respond to all visual information with
The eye and the brain do not respond to all visual information with
same sensitivity.
same sensitivity.
• Some information is neglected during the processing by the
Some information is neglected during the processing by the
brain.because human perception does not involve quantative analysis
brain.because human perception does not involve quantative analysis
of every pixel in the image.
of every pixel in the image.
• Elimination of this information does not affect the interpretation of
Elimination of this information does not affect the interpretation of
the image by the brain.
the image by the brain.
• Psycho visual redundancy is distinctly vision related, and its
Psycho visual redundancy is distinctly vision related, and its
elimination does result in loss of information.
elimination does result in loss of information.
• Quantization is an example. When 256 levels are reduced by grouping
Quantization is an example. When 256 levels are reduced by grouping
to 16 levels, objects are still recognizable.
to 16 levels, objects are still recognizable.
Image Compression Model
Image Compression Model
• Encoder: Compresses the
Encoder: Compresses the
image by reducing
image by reducing
redundancies
redundancies
• Decoder: Reconstructs the
Decoder: Reconstructs the
image from compressed
image from compressed
data
data
BLOCK DIAGRAM OF COMPRESSION MODEL
BLOCK DIAGRAM OF COMPRESSION MODEL
Stages of Encoder
Stages of Encoder
•MAPPER
MAPPER
Reduces Interpixel Redundancy
Reduces Interpixel Redundancy
Reversible operation
Reversible operation
QUANTIZER
QUANTIZER
Reduces Psychovisual Redundancy
Reduces Psychovisual Redundancy
Not a reversible operation
Not a reversible operation
SYMBOL ENCODER
SYMBOL ENCODER
To create a fixed or variable length code
To create a fixed or variable length code
Reversible operation
Reversible operation
/S
LOSSLESS /S
LOSSLESS
Huffman Coding
Huffman Coding
JPEG Data compression
JPEG Data compression
• Joint Photographic Experts Group : lossy compression
Joint Photographic Experts Group : lossy compression
Algorithm of JPEG Data Compression :
Algorithm of JPEG Data Compression :
1.
1.Splitting
Splitting – We split our image into the blocks of 8*8 blocks. It forms
– We split our image into the blocks of 8*8 blocks. It forms
64 blocks in which each block is referred to as 1 pixel.
64 blocks in which each block is referred to as 1 pixel.
2.
2.Color Space Transform
Color Space Transform – In this phase, we convert R, G, B to Y, Cb,
– In this phase, we convert R, G, B to Y, Cb,
Cr model. Here Y is for brightness, Cb is color blueness and Cr
Cr model. Here Y is for brightness, Cb is color blueness and Cr
stands for Color redness. We transform it into chromium colors as
stands for Color redness. We transform it into chromium colors as
these are less sensitive to human eyes thus can be removed.
these are less sensitive to human eyes thus can be removed.
3.
3.Apply DCT
Apply DCT – We apply Direct cosine transform on each block. The
– We apply Direct cosine transform on each block. The
discrete cosine transform (DCT) represents an image as a sum of
discrete cosine transform (DCT) represents an image as a sum of
sinusoids of varying magnitudes and frequencies.
sinusoids of varying magnitudes and frequencies.
4.Quantization
4.Quantization – reduce the no of bit per sample
– reduce the no of bit per sample
5. Serialization –
5. Serialization – In serialization, we perform the zig-zag scanning
In serialization, we perform the zig-zag scanning
pattern to exploit redundancy.
pattern to exploit redundancy.
6. Vectoring
6. Vectoring – We apply DPCM (differential pulse code modeling) on DC
– We apply DPCM (differential pulse code modeling) on DC
elements. DC elements are used to define the strength of colors.
elements. DC elements are used to define the strength of colors.
7.Encoding
7.Encoding –
–
•In the last stage, we apply to encode either run-length encoding or
In the last stage, we apply to encode either run-length encoding or
Huffman encoding. The main aim is to convert the image into text and
Huffman encoding. The main aim is to convert the image into text and
by applying any encoding we convert it into binary form (0, 1) to
by applying any encoding we convert it into binary form (0, 1) to
compress the data.
compress the data.

COMPRESSION MODELSCOMPRESSION MODELSCOMPRESSION MODELS

  • 1.
  • 2.
    WHAT IS COMPRESSION? WHAT IS COMPRESSION ? • Image compression is the process of reducing amount of data Image compression is the process of reducing amount of data required to represent or store an image. required to represent or store an image. • Process of encoding data so that it take less storage space or Process of encoding data so that it take less storage space or less transmission time. less transmission time.
  • 3.
    Why Do WeNeed Image Compression? Why Do We Need Image Compression? • Consider a black and white image that has a resolution of Consider a black and white image that has a resolution of 1000*1000 1000*1000 • each pixel uses 8 bits to represent the intensity. each pixel uses 8 bits to represent the intensity. • So total no of bits required = 1000*1000*8 = 80,00,000 bits per So total no of bits required = 1000*1000*8 = 80,00,000 bits per image. image. • And consider if it is a video with 30 frames per second of the And consider if it is a video with 30 frames per second of the above-mentioned type images above-mentioned type images • total bits for a video of 3 secs is: 3*(30*(8, 000, 000)) total bits for a video of 3 secs is: 3*(30*(8, 000, 000)) • =720, 000, 000 bits =720, 000, 000 bits
  • 4.
    • Storage Efficiency:Compressed images require less storage Storage Efficiency: Compressed images require less storage space.​ space.​ • Transmission Speed: Smaller image files can be transmitted Transmission Speed: Smaller image files can be transmitted faster over networks.​ faster over networks.​ • Cost Reduction: Less storage and bandwidth usage lead to cost Cost Reduction: Less storage and bandwidth usage lead to cost savings.​ savings.​ • Improved Performance: Enhances performance of applications Improved Performance: Enhances performance of applications by reducing load times. by reducing load times.
  • 5.
    Data vs. Information Datavs. Information • Data: Raw pixel values in an image.​ • Information: Meaningful content derived from data.​ • Compression focuses on reducing data redundancy while preserving essential information.
  • 6.
    Redundancy and ItsTypes Redundancy and Its Types •Redundancy means Redundancy means repetitive data or repetitive data or unwanwanted data unwanwanted data CLASSIFICATION CLASSIFICATION  Interpixel Redundancy Interpixel Redundancy Psychovisual Redundancy Psychovisual Redundancy  Coding Redundancy Coding Redundancy
  • 7.
    Coding Redundancy Coding Redundancy Coding redundancy is caused due to poor selection of coding Coding redundancy is caused due to poor selection of coding technique technique  Coding techniques assigns a unique code for all symbols of message Coding techniques assigns a unique code for all symbols of message • Wrong choice of coding technique creates unnecessary additional Wrong choice of coding technique creates unnecessary additional bits. These extra bits are called redundancy bits. These extra bits are called redundancy • CODING REDUNDANCY = AVERAGE BITS USED TO CODE - ENTROPHY CODING REDUNDANCY = AVERAGE BITS USED TO CODE - ENTROPHY
  • 8.
    Interpixel Redundancy Interpixel Redundancy •This type of redundancy is related with the inter-pixel This type of redundancy is related with the inter-pixel correlations within an image. correlations within an image. • The value of any given pixel can be predicted from the value of The value of any given pixel can be predicted from the value of its neighbours or adjacent pixels that are highly corelated. its neighbours or adjacent pixels that are highly corelated. • Inter-pixel dependency is solved by algorithms like: Inter-pixel dependency is solved by algorithms like: • Predictive Coding, Bit Plane Algorithm, Run Length Predictive Coding, Bit Plane Algorithm, Run Length
  • 9.
    Psychovisual Redundancy Psychovisual Redundancy •The eye and the brain do not respond to all visual information with The eye and the brain do not respond to all visual information with same sensitivity. same sensitivity. • Some information is neglected during the processing by the Some information is neglected during the processing by the brain.because human perception does not involve quantative analysis brain.because human perception does not involve quantative analysis of every pixel in the image. of every pixel in the image. • Elimination of this information does not affect the interpretation of Elimination of this information does not affect the interpretation of the image by the brain. the image by the brain. • Psycho visual redundancy is distinctly vision related, and its Psycho visual redundancy is distinctly vision related, and its elimination does result in loss of information. elimination does result in loss of information. • Quantization is an example. When 256 levels are reduced by grouping Quantization is an example. When 256 levels are reduced by grouping to 16 levels, objects are still recognizable. to 16 levels, objects are still recognizable.
  • 10.
    Image Compression Model ImageCompression Model • Encoder: Compresses the Encoder: Compresses the image by reducing image by reducing redundancies redundancies • Decoder: Reconstructs the Decoder: Reconstructs the image from compressed image from compressed data data
  • 11.
    BLOCK DIAGRAM OFCOMPRESSION MODEL BLOCK DIAGRAM OF COMPRESSION MODEL
  • 12.
    Stages of Encoder Stagesof Encoder •MAPPER MAPPER Reduces Interpixel Redundancy Reduces Interpixel Redundancy Reversible operation Reversible operation QUANTIZER QUANTIZER Reduces Psychovisual Redundancy Reduces Psychovisual Redundancy Not a reversible operation Not a reversible operation SYMBOL ENCODER SYMBOL ENCODER To create a fixed or variable length code To create a fixed or variable length code Reversible operation Reversible operation
  • 13.
  • 14.
  • 24.
    JPEG Data compression JPEGData compression • Joint Photographic Experts Group : lossy compression Joint Photographic Experts Group : lossy compression
  • 25.
    Algorithm of JPEGData Compression : Algorithm of JPEG Data Compression : 1. 1.Splitting Splitting – We split our image into the blocks of 8*8 blocks. It forms – We split our image into the blocks of 8*8 blocks. It forms 64 blocks in which each block is referred to as 1 pixel. 64 blocks in which each block is referred to as 1 pixel. 2. 2.Color Space Transform Color Space Transform – In this phase, we convert R, G, B to Y, Cb, – In this phase, we convert R, G, B to Y, Cb, Cr model. Here Y is for brightness, Cb is color blueness and Cr Cr model. Here Y is for brightness, Cb is color blueness and Cr stands for Color redness. We transform it into chromium colors as stands for Color redness. We transform it into chromium colors as these are less sensitive to human eyes thus can be removed. these are less sensitive to human eyes thus can be removed. 3. 3.Apply DCT Apply DCT – We apply Direct cosine transform on each block. The – We apply Direct cosine transform on each block. The discrete cosine transform (DCT) represents an image as a sum of discrete cosine transform (DCT) represents an image as a sum of sinusoids of varying magnitudes and frequencies. sinusoids of varying magnitudes and frequencies.
  • 26.
    4.Quantization 4.Quantization – reducethe no of bit per sample – reduce the no of bit per sample 5. Serialization – 5. Serialization – In serialization, we perform the zig-zag scanning In serialization, we perform the zig-zag scanning pattern to exploit redundancy. pattern to exploit redundancy. 6. Vectoring 6. Vectoring – We apply DPCM (differential pulse code modeling) on DC – We apply DPCM (differential pulse code modeling) on DC elements. DC elements are used to define the strength of colors. elements. DC elements are used to define the strength of colors. 7.Encoding 7.Encoding – – •In the last stage, we apply to encode either run-length encoding or In the last stage, we apply to encode either run-length encoding or Huffman encoding. The main aim is to convert the image into text and Huffman encoding. The main aim is to convert the image into text and by applying any encoding we convert it into binary form (0, 1) to by applying any encoding we convert it into binary form (0, 1) to compress the data. compress the data.