This preview shows pages 1–21. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full DocumentThis preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: 2. Image Compression _+ Motivation — Size of an image = 921,600 bytes
— One second of a Vldeo stream
9 921,600 bytes * 30 frames/sec = 22] Mbit/see
— 90minutes of video _ o 221 Mbits/sec * 5400 seconds = 149 a ham Fundamental Methods + The following signal processing techniques
are used to reduce the size of an image — Subsampling
— Transform Coding
— Entropy Coding
+ Popular Standards: JPEG Subsampling 4 Recall YUV and Y’CbCr schemes which has one brightness component and two color
components + Humans are more sensitive on the changes
in brightness than the changes in colors + Thus, sampling the color components less frequently than the brightness comp ht \
can save the number of pixels ‘ Example of Subsampling 4:' 2:1 heriznntal dummpling, 'nn verti cal dnwsampling I K 4 Y samples for every 2 Ch and 2 Cr 55
samples _
. .1 (f v Subsampling (cont) : reduction in both vertical and horizontal directions Differen'f’ resolutions for luminanee'end chraihin'aﬁ'a’é oOSsible:
 Luminance Yzhigh resolution
 Chrominance U, V: lower resolution Examples: _
 4:2:2: double resolution for luminance Coding
of four
pixels: JPEG  image Preparation Example 4:2:2 YUV, 4:1 :1 YUV, and YUV9 Coding "
 tuminance (Y):  brightness  sampling frequency 13.5 MHz
 Chrominance (U, V): 0 color differences  sampling frequency 6.75 MHz "is System Components 0 Major components of oompruclon m: i'r'ltlnpy
*!‘H mimg (ac) loam CL. Lin, TILE). Discrete Cosine Transform
(DCT) +DCT is the discrete analog of the consine
transform 9 Transformation from spatial to frequency
domain § Redistribute redundancy to enable more
efﬁcient entropy encoding , r” / .‘ + Most current video compression st ds l” x
are DCT—based k . Assumptions:  Data in the transformed domain is easier to compress
 Related processing is feasible " Example:
Fourier Transformation
———————————p
  frequency
time domain domain
41$, ‘_________ bi» Inverse
Fourier Transformation FFT: Fast Fourier Transformation
DCT: Discrete Cosine Transformation General F arm of DC T 0 Mil term 0! DOT for NxN matrix N—l "—1 2 l k I '
at.»  Em.ww[ E E *IWﬂmﬂ—"iFgmﬂ—"h—H n) nlIn1=ﬂ where cm _ (11.5} 1H
kl,krnl,l, : 11...,“ DCT Example 0 Example for 313 image block with range [0.253]: T T
sacral) = %C{kl)c(k2)[n z “n z 'elnl.nz)eue(
I H 1 = nan1 + 1m1 nun2 + nit: T)“‘["T‘)] 98m:  Subtract 128 from each element to center the signal
" around 0
I Perform DCT on original image swung) to yield S(k1,k2)
I Apply quantization matrix, T(k1.k2), on result of DCT to
yield
3(k1ak2)=NlNT(5(k1 sk2VT(k1 #2)),
where NINT is nearest integer function Use of quantization tables for the DCT—coeﬁieients:
 Map.interval of real numbers to one integer number JPEG  Baseline Mode: Entropy COding . . 63 AC coefficients:
 Ordering in ‘zigzag’ form
I A001 A007 DC/’%R//7 A070,,” ‘ ‘ KACT;  reason: coefficients in lower right corner are likely to be zero
 Huffman coding of all coefficients:
 Transformation into a code
a where amount of bits depends on frequency of respective value
%‘  Subsequent runlength coding of zeros DCT Example 0 Example for 313 image block with range [0.253]: T T
sacral) = %C{kl)c(k2)[n z “n z 'elnl.nz)eue(
I H 1 = nan1 + 1m1 nun2 + nit: T)“‘["T‘)] 98m:  Subtract 128 from each element to center the signal
" around 0
I Perform DCT on original image swung) to yield S(k1,k2)
I Apply quantization matrix, T(k1.k2), on result of DCT to
yield
3(k1ak2)=NlNT(5(k1 sk2VT(k1 #2)),
where NINT is nearest integer function (1) Original Frame I“WI"'1’" 111111 11131311113131
mmmmmmm
raummmmm 0“ 11111113111151.11
mmmnnnnumm
I'HIIIIIHIIJIIIII
mmmmmmm
mmmmmmm (3) Mar DOT Suﬁ*1}: 313 51 4111111111.
4341 1311314444
411411113311411114
111—11 1 11 11—11—13 1
—i 1 1 4—3—1 —3 3
1 3 13114 11 3
4 :1 4—1—11 1 4
1 1 14—1: 3 1 DC T Example ( cont. ) (2) Subtract "128" ""rW' 531—3415113343": 55154145553514!
5144' 43 54514533!
454'! 5145513133!
545' 51 455435145
5151 51 515541441
5151 51 545541145
5151 53 515541141 (4) Quantization Matrix 1111.11; . 15111111514 411 51 51
1111141515 55 ill 55
1413151444 51' 55 55
14111115 51 5? BI 51
1511315555 11191031?
14.355554511114113 51
“54155111111113”!
115155551111I1I3” DC T Example (cont. ) (5) After Dlvlnlon by Mutilation matrix .... “kll)
Suﬁ"1} ' 71W“ I I 413211
3*J 1 I] I I
4] l n I I
lllll
IIIII
IIIII
IIIII
IIIII H (6) 219—239 Scan —I 3531 43111 400 1'1r'9r' cannon“ a 10000
ooau0ﬁ°°°°°°ﬂ ‘ﬁﬁunﬂﬁ Oi i H uﬁcman Encoding (cont) 0 Symbols with higher probabilities are assigned shorter
codewords. Symbol Code Probability 32,32
31 0 p1: 5f3 "" 1%; I 32 100 [32: 3’32 " '
53 110 [33: W32 '
34: 1110 p4: 1332  .
55 101 p5=1i&=%z ss 1 111 p3: 1332 ' I Unliormlength code:
I Huffman code: I Optimal (entropy): 5. Entropy Coding: Assumption:
 Long sequences of identical symbols Example: ...ABCEEEEEEDACB... compression y: DACB... symbol number of
occurrences special flag Special variant: zerolength encoding
 onlv repetition of zeroes count J PEG 0Became an ISO international standard in
1992 + Use most of the techniques introduced
earlier § Both. of the sequential and progressive
presentations are supported I / ' I ...__...._\~ I u
$0“
a '15 1:; ﬁ‘
' a
i m. I;
(E
J (IraFf" ‘1 K N. as“ ~. wee w" ' a):
.9352” g
Wié'mwmﬁm.‘ 2 x JPEG v15 '25 5:4 it“: Very general compression scheme 'Independence of:
 Image resolution
 Image and pixel aspect ratio
 Color representation
 Image complexity and statistical characteristics Welldefined Interchange format of encoded data Implementation in:
 Software only
 Software and hardware “MOTION JPEG” for video compression
 Sequence of JPEGencoded images ...
View
Full
Document
This note was uploaded on 11/30/2011 for the course CIS 6930 taught by Professor Staff during the Fall '08 term at University of Florida.
 Fall '08
 Staff

Click to edit the document details