Image Storage and Compression

Download Report

Transcript Image Storage and Compression

Slide 1

‫این دو عالم علم دارد در نهاد منتخب‬
‫وان جهانی رمز دارد در حروف مختصر‬
‫سنایی غزنوی‬


Slide 2

‫‪Image/Video Compression & VOD‬‬
‫برنا فیروزي‬
‫نوید زرین درفش‬
‫محمود قدیمي‬
‫مهندسي فناوري اطالعات‬
‫بهار ‪88‬‬


Slide 3

Image Compression

3


Slide 4

‫‪Compression‬‬
‫‪‬‬

‫‪‬‬
‫‪‬‬
‫‪‬‬

‫فشرده سازی (‪ ،)Compression‬پردازشی است که با حذف‬
‫اطالعات اضافی‪ ،‬داده ها را به عالیم دیجیتالی کاهش می دهد‪.‬‬
‫این پردازش بسته به پهنای باند مورد نیاز برای انتقال داده ها و‬
‫میزان فضای ذخیره سازی‪ ،‬داده ها را کاهش می دهد‪.‬‬
‫کاهش پهنای باند مورد نیاز امکان انتقال داده های بیشتری را در‬
‫یک زمان واحد می دهد‪.‬‬
‫كاهش افزونگی (‪)redundancy‬‬
‫‪4‬‬


Slide 5

‫ديجيتال ‪،‬آنالوگ‬
‫‪‬‬

‫پردازش تصاویر در فضاي دیجیتال انجام مي شود‬

‫‪‬‬

‫اگر ما یك منبع تصویري آنالوگ داشته باشیم باید قبل از انجام هر‬
‫عمل فشرده سازي ابتدا به آن را به دیجیتال تبدیل كنیم‬

‫‪5‬‬


Slide 6

Purpose of Image Compression







6

Saving storage space
Saving transfer time
Easy processing
Easy to transmitted over network
reduce cost


Slide 7

‫هدف‬
‫‪‬‬

‫در حالت ایده آل ما خواستار‬
‫‪‬‬
‫‪‬‬

‫‪‬‬
‫‪‬‬

‫حداكثر كیفیت تصویر‬
‫حداقل فضاي ذخیره سازي و پردازش منابع‬

‫ما نمي توانیم هر دو هدف را در بهترین شرایط داشته باشیم‬
‫بهترین فشرده سازي چگونه است؟‬

‫‪7‬‬


Slide 8

Why we want to compress?
To transmit an RGB 512X512, 24 bit image
via modem 28.2 kbaud(kilobits/second)
( 512  512 pixels )( 24 bits / pixel )
( 28 . 8  1024 bits / second )

8

 213 second

(  4 min)


Slide 9

Image Compression Coding

lossless
compression
Image
Compression
lossy
compression

9

Huffman
Coding

Run-length
encoding
Predictive
Coding
Transform
Coding


Slide 10

‫دو كالس اصلي فشرده سازي تصاوير‬
‫‪‬‬
‫‪‬‬

‫‪‬‬

‫‪‬‬

‫‪ )Lossy‬پراتالف(‬
‫از روي دیتاهاي تصویر ذخیره‬
‫شده مي توان به تصویري نزديك‬
‫به تصویر اصلي رسید‬
‫تصویر ذخیره شده خود تصویر‬
‫اصلی نیست‪ ،‬بلکه شبیه آن است و‬
‫اطالعاتی را از دست داده است‬
‫‪Compression rate:‬‬
‫‪high compression‬‬

‫‪‬‬
‫‪‬‬

‫‪‬‬

‫‪‬‬

‫‪( Lossless‬بدون اتالف(‬
‫از روي دیتاهاي تصویر ذخیره‬
‫شده مي توان دقیقا به تصوير‬
‫اصلي رسید‬
‫تصویر ذخیره شده بدون از دست‬
‫دادن كمترین داده ای‪ ،‬خود تصویر‬
‫است‪.‬‬
‫‪Compression rate:‬‬
‫)‪2:1 (at most 3:1‬‬

‫‪10‬‬


Slide 11

General compression system model

O utput ima ge

Input im a ge

Source decoder

Source encoder

D ecoder

E ncoder

C hannel decoder

C hannel encoder

C hannel

11


Slide 12

Compression System Model


Compression

Input

Preprocessing

Encoding

Compressed
File

• Decompression
Compressed
File

12

Decoding

Postprocessing

Output


Slide 13

Compression Ratio
Compressio n Ratio 

Uncompress ed File Size
Compressed

File Size

 CR

Ex Image 256X256 pixels, 256 level grayscale
can be compressed file size 6554 byte.
Original Image Size = 256X256(pixels) X
1(byte/pixel)
= 65536 bytes
compressio n Ratio 

13

65536
6554

 10


Slide 14

Bits per Pixel
(2) Bits per Pixel  Number of Bits
Number of Pixels

Ex Image 256X256 pixels, 256 level grayscale
can be compressed file size 6554 byte.
Original Image Size = 256X256(pixels) X
1(byte/pixel)
= 65536 bytes
Compressed file
= 6554(bytes)X8(bits/pixel)
= 52432 bits
14

Bits per Pixel 

52432
65536

 0 .8


Slide 15

Key of compression


Reducing Data but Retaining Information

Various amounts of data can be used to represent the same
amount of information. It’s “Data redundancy”

Relative data redundancy
RD  1 

15

1
CR


Slide 16

Entropy


Average information in an image.
L 1

Entropy    p k log 2 ( p k )
k 0

• Average number of bits per pixel
L 1

La 

l

k

pk

k 0

pk 

16

nk
n

, where k  0 ,1,  , L  1


Slide 17

Compression Standard
Standard:

17

Image

Video

ISO

JPEG
JPEG2000

MPEG1,MPEG2,
MPEG4, MPEG7

ITU

N/A

H.261, H.263, H.263+,
H.26L


Slide 18

Image Compression Coding

lossless
compression
Image
Compression
lossy
compression

18

Huffman
Coding

Run-length
encoding
Predictive
Coding
Transform
Coding


Slide 19

Loseless Compression





19

No data are lost
Can recreated exactly original image
Often the achievable compression is mush less


Slide 20

Huffman Coding ( VLC,Entropy)



Using Histogram probability
5 Steps
1.
2.
3.
4.
5.

20

Find the histogram probabilities
Order the input probabilities(smalllarge)
Addition the 2 smallest
Repeat step 2&3, until 2 probability are left
Backward along the tree assign 0 and 1


Slide 21

Huffman Coding(cont)


40
30
20
10

Step 1 Histogram Probability
p0 = 20/100 = 0.2
p1 = 30/100 = 0.3
p2 = 10/100 = 0.1
p3 = 40/100 = 0.4

0

1

2

3

 Step 2 Order
p3  0.4
p1  0.3
p0  0.2
p2  0.1


Slide 22

Huffman Coding(cont)
 Step 3,4 Add 2 smallest
p 3  0 .4  0 .4  0 .4
p1  0 . 3  0 . 3
0 .6
p 0  0 .2
0 .3

0 .6
0 .4

p 2  0 .1

 Step 5 assign 0 and 1

22

Natural Code
00

Probability
0.2

Huffman Code
010

01

0.3

00

10

0.1

011

11

0.4

1


Slide 23

Huffman Coding(cont)




The original Image :average 2 bits/pixel
The Huffman Code:average
3

La 

l

i

p i  3 ( 0 . 2 )  2 ( 0 . 3 )  3 ( 0 . 1)  1( 0 . 4 )  1 . 9

i0

3

Entropy    p i log 2 ( p i )
i0

 ( 0 . 2 ) log 2 ( 0 . 2 )  ( 0 . 3 ) log 2 ( 0 . 3 ) 
 


(
0
.
1
)
log
(
0
.
1
)

(
0
.
4
)
log
(
0
.
4
)
2
2



23

 1 . 846 bits / pixel


Slide 24

Run Length Encoding


24

Run-length encoding (RLE) is a very simple form
of data compression in which runs of data (that is,
sequences in which the same data value occurs in
many consecutive data elements) are stored as a
single data value and count, rather than as the
original run. This is most useful on data that
contains many such runs: for example, relatively
simple graphic images such as
icons, line drawings, and animations.


Slide 25

Run-Length Coding





25

Counting the number of adjacent pixels with
the same gray-level value
Used primarily for binary image
Mostly use horizontal RLC


Slide 26

Run-Length Coding(cont)
Binary Image 8X8

26

horizontal

0

0

0

0

0

0

0

0

1st Row

8

1

1

1

1

0

0

0

0

2nd Row

4,4

0

1

1

0

0

0

0

0

3rd Row

1,2,5

0

1

1

1

1

1

0

0

4th Row

1,5,2

0

1

1

1

0

0

1

0

5th Row

1,3,2,1,1

0

0

1

0

0

1

1

0

6th Row

2,1,2,2,1

1

1

1

1

0

1

0

0

7th Row

4,1,1,2

0

0

0

0

0

0

0

0

8th Row

8


Slide 27

Example of RLE


Let us take a hypothetical single scan line, with B
representing a black pixel and W representing white:

WWWWWWWWWWWWBWWWWWWWWWWWWBBBWWWW
WWWWWWWWWWWWWWWWWWWWBWWWWWWWWWW
WWWW
If we apply the run-length encoding (RLE) data compression
algorithm to the above hypothetical scan line, we get the
following:12W1B12W3B24W1B14W

27


Slide 28

Lossy Compression






28

Allow a loss in the actual image data
Can not recreated exactly original image
Commonly the achievable compression is
mush more
Such as JPEG


Slide 29

Predictive Coding
Common Predictive
Coding

29

DM (Delta Modulation)

DPCM (Differential Pulse
Code Modulation)


Slide 30

The principle of Predictive Coding


30

The system consists of an encoder and a decoder,
each containing an identical predictor. As each
successive pixel of the input image, is introduced to
the encoder, the predictor generates the anticipated
value of that pixel based on some number of past
inputs. The output of the predictor is then rounded
to the nearest integer.


Slide 31

Predictive coding model I
Input image

n

fn

Symbol
encoder

fˆn
Predictor
Compressed image

Output image

fn'

Predictor

31

‘
fˆ n '

Symbol
decoder


Slide 32

Predictive coding model II
The predictor :

fˆ n  F ( f n  1 , f n  2 ,  , f n  k )


n

 f n  fˆ n

The symbol encoder : generate the next element of the
compressed data stream
Decoder : perform the inverse of encoding
The linearity predictor :

32

fˆ n  F ( f n  1 , f n  2 ,  , f n  k ) 

n 1

a
kl

k

fk ,

a

k

1


Slide 33

Delta Modulation I
Delta modulation (DM or Δ-modulation) is an analog-todigital and digital-to-analog signal conversion technique
used for transmission of voice information where quality
is not of primary importance.
DM is the simplest form of differential pulse-code
modulation (DPCM) where the difference between
successive samples is encoded into n-bit data streams. In
delta modulation, the transmitted data is reduced to a 1bit data stream.

33


Slide 34

Differential Pulse Code Modulation


34

Differential Pulse Code Modulation (DPCM) compares
two successive analog amplitude values, quantizes and
encodes the difference, and transmits the differential
value.


Slide 35

Transform Coding I (DCT)




35

Transform coding is a type of data compression for
"natural" data like audio signals or photographic
images.
The transformation is typically lossy, resulting in a
lower quality copy of the original input.


Slide 36

Transform Coding II

36


Slide 37

Transform Coding III
A transform coding system
Input image

Construct
subimages

Compressed
image

37

Quantizer

Forward
transform

Symbol
encoder
Compressed
image

Symbol
decoder

Inverse
transform

Merge
subimages


Slide 38

BMP (Bitmap) - lossless









38

Use 3 bytes per pixel, one
each for R, G, and B
Can represent up to 224 =
16.7 million colors
No entropy coding
File size in bytes =
3*length*height, which can be
very large
Can use fewer than 8 bits per
color, but you need to store
the color palette
Performs well with ZIP, RAR,
etc.


Slide 39

GIF (Graphics Interchange Format)


Can use up to 256 colors from
24-bit RGB color space




39

If source image contains
more than 256 colors, need
to reprocess image to fewer
colors

Suitable for simpler images
such as logos and textual
graphics, not so much for
photographs


Slide 40

JPEG (Joint Photographic Experts
Group) - lossly





40

Most dominant image
format today
Typical file size is about
10% of that of BMP (can
vary depending on
quality settings)
Unlike GIF, JPEG is
suitable for photographs,
not so much for logos
and textual graphics


Slide 41

Joint Picture Expert Group


41

The name "JPEG" stands for Joint Photographic Experts Group,
the name of the committee that created the standard. The group
was organized in 1986, issuing a standard in 1992, which was
approved in 1994 as ISO 10918-1. JPEG is a commonly used
method of compression for photographic images. The degree of
compression can be adjusted, allowing a selectable tradeoff
between storage size and image quality.


Slide 42

JPEG2000


42

JPEG 2000 is a wavelet-based image compression standard. It
was created by the Joint Photographic Experts Group committee
in the year 2000 with the intention of superseding their original
discrete cosine transform-based JPEG standard (created about
1991).


Slide 43

JPEG Encoding Steps

43



Preprocess image



Apply 2D forward DCT



Quantize DCT coefficients



Apply RLE, then entropy encoding


Slide 44

JPEG Block Diagram
8x8 blocks

Source
Image
B
G

R
DCT-based encoding
FDCT

44

Quantizer

Entropy
Encoder

Table

Table

Compressed
image data


Slide 45

Image Compression- JPEG


Using the DCT, the entries in Y will be organized
based on the human visual system.



The most important values to
our eyes will be placed in the
upper left corner of the matrix.



The least important values
will be mostly in the lower
right corner of the matrix.

Most
Important
SemiImportant
Least
Important

45


Slide 46

JPEG
8 x 8 Pixels

46

Image


Slide 47

Image Compression



Gray-Scale Example:
Value Range 0 (black) --- 255 (white)

63 33 36 28 63 81
27 18 17 11 22 48
72 52 28 15 17 16
132 100 56 19 10 9
187 186 166 88 13 34
184 203 199 177 82 44
211 214 208 198 134 52
211 210 203 191 133 79

47

86 98
104 108
47 77
21 55
43 51
97 73
78 83
74 86


Slide 48

Image Compression


2D-DCT of matrix

-304 210
-327 -260
93 -84
89 33
-9 42
-5 15
10
3
12 30

48

104
67
-66
-19
18
-10
-12
0

-69
70
16
-20
27
17
-1
-3

10
-10
24
-26
-7
32
2
-3

20
-15
-2
21
-17
-15
3
-6

-12 7
21 8
-5 9
-3 0
29 -7
-4 7
-2 -3
12 -1


Slide 49

Image Compression
-304 210
-327 -260
93 -84
89 33
-9 42
-5 15
10
0
0
0

104
67
-66
-19
18
0
0
0

-69
70
16
-20
0
0
0
0

10 20 -12 0
-10 -15 0 0
24 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0

As you can see, we save a little over half the original
memory.

49


Slide 50

Reconstructing the Image


New Matrix and Compressed Image

55 41 27 39 56 69 92 106
35 22 7 16 35 59 88 101
65 49 21 5 6 28 62 73
130 114 75 28 -7 -1 33 46
180 175 148 95 33 16 45 59
200 206 203 165 92 55 71 82
205 207 214 193 121 70 75 83
214 205 209 196 129 75 78 85

50


Slide 51

Can You Tell the Difference?
Original

51

Compressed


Slide 52

Image Compression
Original

52

Compressed


Slide 53

Example - One everyday photo with
file size of 2.76 MB

53


Slide 54

Example - One everyday photo with
file size of 600 KB

54


Slide 55

Example - One everyday photo with
file size of 350 KB

55


Slide 56

Example - One everyday photo with
file size of 240 KB

56


Slide 57

Example - One everyday photo with
file size of 144 KB

57


Slide 58

Example - One everyday photo with
file size of 88 KB

58


Slide 59

Analysis




Near perfect image at 2.76M, so-so image at 88K
Sharpness decreases as file size decreases
Which file size is the best?



59

No correct answer to this question
Answer depends upon how strict we are about image quality, what
purpose image is to be used for, and the resources available


Slide 60

Conclusion





60

Image compression is important
Image compression has come a long way
Image compression is nearly mature, but there
is always room for improvement


Slide 61

Video Compression


Slide 62

Video Data Size
size of uncompressed video in gigabytes
1 sec
1 min
1 hour
1000 hours

1920x1080
0.19
11.20
671.85
671,846.40

1280x720
0.08
4.98
298.60
298,598.40

640x480
0.03
1.66
99.53
99,532.80

320x240
0.01
0.41
24.88
24,883.20

160x120
0.00
0.10
6.22
6,220.80

image size of video

1280x720 (1.77)

62

640x480 (1.33)

320x240

160x120


Slide 63

Video Bit Rate Calculation
width * height * depth * fps
= bits/sec

compression factor
width ~ pixels(160, 320, 640, 720, 1280, 1920, …(
height ~ pixels
(120, 240, 480, 485, 720, 1080, …(
depth ~ bits (1, 4, 8, 15, 16, 24, …(
fps ~ frames per second (5, 15, 20, 24, 30, …(
compression factor (1, 6, 24, …(

63


Slide 64

Effects of Compression
storage for 1 hour of compressed video in megabytes
1:1
3:1
6:1
25:1
100:1

1920x1080
671,846
223,949
111,974
26,874
6,718

1280x720
298,598
99,533
49,766
11,944
2,986

640x480
99,533
33,178
16,589
3,981
995

3 bytes/pixel, 30 frames/sec

64

320x240
24,883
8,294
4,147
995
249

160x120
6,221
2,074
1,037
249
62


Slide 65

Channel Bandwidths

65


Slide 66

Channel Bandwidth

66


Slide 67

Application Requirements

67


Slide 68

Source Video Formats

68


Slide 69

The Need for Video Compression


High-Definition Television (HDTV)
1920x1080
– 30 frames per second (full motion)
– 8 bits for each three primary colors (RGB)
Total 1.5 Gb/sec!




Cable TV: each cable channel is 6 MHz
Max data rate of 19.2 Mb/sec
– Reduced to 18 Mb/sec w/audio + control …
Compression rate must be ~ 80:1!


69


Slide 70

The Need for Video Compression


Some figures


Uncompressed video -> big amount of data






Communication and storage capacities limits




70

Color picture 800x320 pix, 24 bits/pix -> 6.3 Mbit/s
SDTV 720x480, 30Hz, 16 bits/pix
-> 166 Mbit/s
HDTV 1920x1080, 30Hz, 16 bits/pix -> 1Gbit/s
Cable or satellite bandwidth : 38 Mbit/s
ADSL : 1 to 8 Mbit/s
DVD capacity : 5 to 8 GB


Slide 71

‫كاربرد‬


Video compression is now everywhere :






71

TV broadcasting over cable, satellite or terrestrial
networks,
CD-ROM, DVD, PC video storage,
Videophone and teleconferencing,
(VoD, IPTV),
Video over moInternet streaming biles.


Slide 72

Standardization (1/2)
Video codecs
Technologies

1950’s

1960’s

1970’s

1980’s

DPCM 52/80
MC Prediction 72/89
Transform Coding 65/80

Wavelet 85/-H.261

CD-ROM 1-1.5Mb/s
Digital TV, DVD

25 to 50 Mb/s

Videophone
Video streaming & post-prod

MPEG-1

4 to 80 Mb/s

Camcorder, VTR

Standards

2000’s

Entropy Coding 49/76

Videophone 56Kb/s – 2Mb/s

72

1990’s

30 Kb/s

30 Kb/s to 600Mb/s

H.262/MPEG62
DVCPRO

H.263
MPEG4 ASP
H.264/AVC

Convergence of all video applications, digital cinema
30 Kb/s to 600Mb/s

SMPTE/VC1
H.264/SVC


Slide 73

 Bit rate evolution for SDTV Broadcast

Bit rate evolution

C. Ratio from 4:2:2

6

28

1st generation encoders
2nd generation encoders
(Stat-Mux + Rate control improvements)

5

3rd generation encoders
(advanced Pre-processing)

Mbit/s

4

3

1st generation
encoders

2

2nd generation
encoders

1

166

0

73

1995

1997

1999

2001

2003

2005

2007

2009


Slide 74

MPEG Compression



Compression through



74

Spatial
Temporal


Slide 75

Video Redundancies




75

Spatial
Neighboring pixels in a frame are statistically
related.
Temporal
Pixels in consecutive frames are statistically
related.
One can achieve higher compression ratios
by exploiting both spatial and temporal
redundancies


Slide 76

Spatial Redundancy


76

Take advantage of similarity among most
neighboring pixels


Slide 77

Spatial Redundancy Reduction









77

RGB to YUV
– less information required for YUV (humans less
sensitive to chrominance)
Macro Blocks
– Take groups of pixels (16x16)
Discrete Cosine Transformation (DCT)
– Based on Fourier analysis where represent signal as
sum of sine's and cosine’s
– Concentrates on higher-frequency values
– Represent pixels in blocks with fewer numbers
Quantization
– Reduce data required for co-efficients
Entropy coding
– Compress


Slide 78

Motion Compensation


Macro Block



Motion Vector
1 2 3
4 5 6
7 8 9

78

16 x 16

1 2 3
4 5 6
7 8 9


Slide 79

Video compression in MPEG-1&2


Spatial redundancy reduction (DCT example)

139
144
150
159
159
161
162
162

144
151
155
161
160
161
162
162

149
153
160
162
161
161
161
161

153
156
163
160
162
161
163
161

155
159
158
160
162
160
162
163

155
156
156
159
155
157
157
158

155
156
156
159
155
157
157
158

155
156
156
159
155
157
157
158

DC T

1260 -1 -12
-23 -17 -6
-11 -9 -2
-7 -2
0
-1 -1
1
2
0
2
-1
0
0
-3
2 -4

-5
-3
2
1
2
0
-1
-2

2
-3
0
1
0
-1
0
2

-2
0
-1
0
-1
1
2
1

-3
0
-1
0
1
1
1
-1

Q uan tsi a toi n
158
-1
-1
0
0
0
0
0

79

0
-1
0
0
0
0
0
0

-1
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

0
0
0
0
0
0
0
0

z gi -zag scan
158 0 -1 -1 -1 -1 EO B

1
-1
0
0
1
-1
-1
0


Slide 80

Spatial Redundancy Reduction

“Intra-Frame
Encoded”
Quantization
• major reduction
• controls ‘quality’

80

Zig-Zag Scan,
Run-length
coding


Slide 81

Loss of Resolution
Original (63 kb)

Low (7kb)

Very Low (4 kb)

81


Slide 82

Temporal Redundancy


82

Take advantage of similarity between
successive frames

950

951

952


Slide 83

Temporal Redundancy Reduction

83


Slide 84

Temporal Redundancy Reduction

84


Slide 85

Temporal Redundancy Reduction





frames are independently encoded
P frames are based on previous I, P frames
– Can send motion vector plus changes
B frames are based on previous and following I and P
frames
– In case something is uncovered


Slide 86

Group of Pictures (GOP)






Starts with an I-frame
Ends with frame right before next I-frame
“Open” ends in B-frame, “Closed” in P-frame
MPEG Encoding a parameter, but ‘typical’:



86

IBBPBBPBBI
IBBPBBPBBPBBI


Slide 87

Question



87

When may temporal redundancy reduction be
ineffective?


Slide 88

Answer



When may temporal redundancy reduction be
ineffective?



88

Many scene changes
High motion


Slide 89

Non-Temporal Redundancy


89

Sometimes high motion


Slide 90

Typical Compress. Performance

90

Type Size Compression
--------------------I
18 KB
7:1
P
6 KB
20:1
B
2.5 KB
50:1
Avg 4.8 KB
27:1
--------------------Note, results are Variable Bit Rate, even if
frame rate is constant


Slide 91

Inter and Intra coding
To exploit spatial redundancies within a
frame (Intra coding):
8x8 DCT, similar to JPEG
 To exploit temporal redundancies between
frame (Inter coding):
Motion Estimation


91


Slide 92

Frame Types

Two frame types:



92

Intra-frames (I-frames): I-frame provides an
accessing point, it uses basically JPEG.
Inter-frames (P-frames): P-frames use from previous
frame ("predicted"), so frames depend on eachother.


Slide 93

Some video formats


The 4:2:2 format











The 4:2:0 format


93

Y sampling @ 13.5 MHz
C sampling @ 6.75 MHz
8 bits per pixel
720 active points per line
576 lines active lines per image (2 fields) (625 lines)
and 480 active lines (525 lines)
Pixels are not square (e.g. for 480 lines, only 640 active points
are needed - VGA format)
Image size 720*576 or 720*480
Vertical chrominance resolution reduced by a factor 2
(average on two successive lines)


Slide 94

Some video formats


SIF format (Source Intermediate Format)
Half the vertical & horizontal resolution of 4:2:0
For 50Hz countries:





CIF format (Common Intermediate Format)






Intermediate format used in videoconferencing
(communication between US & Europe)
resolution: 360*288
Sampling frequency: 30 Hz

QCIF (Quarter CIF)


94

Luminance: 360*288
Chrominance: 180*120

Half the vertical & horizontal resolution of CIF.

AV Compression / Alain Bouffioux

December, 20, 2006


Slide 95

MPEG




95

MPEG was an early standard for lossy compression
of video and audio.
Development of the MPEG standard began in May
1988.


Slide 96

MPEG Coding Performance


Decoding is easy






Encoding is expensive






96

MPEG1 decoding in software on most platforms
Hardware decoders are widely available ($150/board)
Windows graphics accelerators with MPEG decoding
now entering market (e.g., Matrox, Diamond, …(
Sequential software encoders are 20:1 real-time
Real-time encoders use parallel processing
Real-time hardware encoders are expensive (e.g.,
$12K-$50K for MPEG1 and $100K-$500K for MPEG2)
Hardware-assisted off-line MPEG1 encoders (3:1) used
for multimedia authoring at reasonable cost ($5k)


Slide 97

MPEG







MPEG1: low bitrate
MPEG2: VCD, DVD
MP3: MPEG2 profile 3, for music
MPEG4: Network streaming
MPEG7: Searching and indexing


Slide 98

MPEG Standards


MPEG-1 ~ 1-1.5Mbps (early 90s) - vhs quality
(1992)





98

Frame encoding
For compression of 320x240 full-motion video at rates
around 1.15Mb/s
Applications: video storage (VCD)
CIF images, 4:2:0 sampling, 1.5 Mbs


Slide 99

MPEG Standards


MPEG-2 ~ 2-80Mbps (mid 90s) - broadcast quality
(1994)









99

Frame and field encoding
For higher resolutions
Support interlaced video formats and a number of features
for HDTV
Address scalable video coding
Also used in DVD
CCIR 601 images, 4:2:2 sampling, 15 Mbs
Interlaced and progressive scanning


Slide 100

MPEG Today




100

MPEG4 ~ 9-40kbps (later 90s)


Around Objects, not Frames



For very low bit rate video and audio coding



Applications: interactive multimedia and video telephony



Lower bandwidth

MPEG-7


Provide a fast and efficient searching, filtering



New standard



Internet orientated



VOP (Video Object Plane)



Profiles and levels


Slide 101

H.26x







H.261: first generation
H.262 (MPEG2)
H.263: video conference
H.263++: enhancement
H.26L: latest


Slide 102

VOD

Watch
what you want
when you want


Slide 103

Standardization (2/2)
DVB Transport

DVB-S

DVB-S2

DVB-C

Satellite TV

Cable

DVB-T

Terrestrial TV
DVB-H

Mobile TV

DVB-IPI in progress

103

1995

1997

1999

2001

2003

2005

2007

IPTV
2009


Slide 104

Video on Demand





One video server
Many video data
Many clients
Client want to watch at any time

Clients can send request to the server and request to
watch a particular video. The server have to
response by streaming the requested video to the
client. We want to be able to support large
number of clients, and clients should be able to
watch at anytime.
104


Slide 105

Streaming


Streaming media:
the ‘real-time’ playing of a video-, audioand/or datastream on a machine from the
moment the first bytes come in.



VoD such as YouTube, MSN Video, Google
Video, Yahoo Video, CNN…
2006 April to December: MSN Video service
Client-server mode
Covering over 520 million streaming requests for
more than 59,000 videos.



105


Slide 106

Streaming: live vs on-demand

unicast

multicast

archive

realtime

VoD

Eventdriven

scheduled

Eventdriven


Slide 107

Streaming: live vs on-demand






107

Live audio and/or video streaming is a completely
different sport from on-demand streaming.
VoD (Video on demand) is unicast streaming from
an archive.
Scheduled (a tv like netcast) is usually multicast
streaming from an archive
Live (realtime) streaming can be done both in
unicast and multicast.


Slide 108

Motivation




108

VoD such as YouTube, MSN Video, Google
Video, Yahoo Video, CNN…
As the trend of increasing demands on such
services and higher-quality videos, it becomes
a costly service to provide.


Slide 109

Video-on-Demand Distribution
Model
M u lt ic a s t / B r o a d c a s t C a p a b le
N etw ork
STB

STB

STB

VID EO
SERVER




109

A client can tune in to receive any ongoing
media delivery using its Set Top Box
True broadcast: Satellite and cable TV
networks
109


Slide 110

Rewind TV


VoD servers support rewind function

VoD servers
customer premise
TV
IP backbone

Unicast

Set-top box

110


Slide 111

IPTV/VOD : Full Operator model
Content Provider

Service Provider

CPE

Network Provider
DSL (Copper)

IP Streamer

Broadcast
feed

DSLAM
RT Encoder

Encryption
Server

Optical Fiber

VOD Server
Encoder

Media
database

ADSL
Modem/Router

Content
Management
Server

Core &
Aggregation
Networks

STB
Optical
Line
Terminatio
n

Optical Network
Termination

HFC (Coax)
Licence
Server

EPG
metadata

111

Application
Servers

Remote
Management
Server
111

Subscriber
Management
and Billing
Server

Cable
Modem
Termination
System

Cable
Modem/Router


Slide 112

What is IPTV?


The fundamentals


IPTV = Internet Protocol Television






112

Digital TV service delivered over a broadband network
using the Internet Protoc0l

IPTV usually refers to TV services over a
Network Operator’s quality controlled network.
Internet TV = IPTV over the public Internet


Slide 113

IPTV Network
 One of the first

 One of the largest
 150 TV channels
 250,000 subscribers

113


Slide 114

Basic IPTV Structure

114


Slide 115

Set Top Box – Home gateway (BST)

115


Slide 116

IPTV Set-Top Box

Broadcast TV

116

DVR

Movies On Demand

On Demand

116


Slide 117

Other Services


VOD – Video on Demand




DVR




117

Record Live TV from your set top box

HD




Watch what you want when you want

High Definition TV. The evolution of Television

On line Gaming


Slide 118

Logout | Help

Search

My Content

Annotate

Go
Title: The Hidden Child
Description: Of the 1,600,000 Jewish

Selected Video

children who lived in Europe before
World War II, only 100,000 survived
the Holocaust.

Title: Ralph Golzio interview on
Paterson Silk Strike of 1913
Description: Ralph Golzio was a
teenager at the time of the Paterson
Silk Strike of 1913.

>>

<<

Title: Sondra Gash interview
Description: Evelyn Hershey of the
American Labor Museum interviews

Search For and Select a Video for
Annotation
118

--


Slide 119

Logout | Help

Selection of
Search
My Content
video found
through search

Annotate

Go
Title: The Hidden Child
Description: Of the 1,600,000 Jewish
children who lived in Europe before
World War II, only 100,000 survived
the Holocaust.

Title: Ralph Golzio interview on
Paterson Silk Strike of 1913
Description: Ralph Golzio was a

-

teenager at the time of the Paterson
Silk Strike of 1913.

>>

<<

Title: Sondra Gash interview
Description: Evelyn Hershey of the
American Labor Museum interviews

Drag and Drop Thumbnails Here To Create
an Annotation
119

--


Slide 120

Logout | Help

Once an annotation is
saved it will appear in the
videos timeline.

Amsterdam, The
Netherlands

Search
Start Time:

My Content
5:01

End Time:

8:34

Title:

Amsterdam, The Netherlands

Note:

City street in Amsterdam.

-

Private

My Institution Only

Indexed/Searchable:
>>

Annotate



NJVid Wide

Save

<<

Start and stop
markers can be
placed.

Metadata
placed here.

Annotate and Capture Start/End Times
120


Slide 121