Image Compression : Basic Concept

1
Image Compression Basic Concept
2

• Image is accessed (??) as a 2-D array (??) of
  data, where each data point is referred to as a
  pixel (??)
• Notation
• I(r,c) Brightness (??) of image at the pt
  (r,c)
• where
• r row(?), and c column(?)

3
Visible Light Imaging

• Reflectance (??) function determines manner in
• which objects (??) reflect light

4

• Sensors Converts (??) light energy into
  electrical energy

a) Single imaging sensor b) Linear ( line)
sensor c) 2-D or array sensor

• CCD 4kx4k CMOS less power, cheaper, image
  quality not as good as CCD

5
Image Representation

• Optical (??) image Collection of spatially
  distributed (????) light energy measured by an
  image sensor to generate I(r,c)
• Matrix 2-D array like the image model,
• I(r,c)
• Vector One row or column in a matrix

6
Image Types

• Binary (???) images Simplest type of images,
  which can take two values, typically black or
  white, or 0 or 1
• Gray scale (??) images One-color or monochrome
  images that contains only brightness information
  and no color information
• Color images 3 band monochrome images, where
  each band corresponds to a different color,
  typically red, blue and green or RGB

7

• Color pixel vector Single pixels values for a
  color image, (R,G,B)
• Multispectral(???) Images Images of many bands
  containing information outside of the visible
  spectrum(????)

8
Color Transform/Color Model

• Mathematical model or algorithm to map(??) RGB
  data into another color space (????)
• Decouples (??) brightness and color information
• Hue(??)/Saturation(???)/Lightness(??) (HSL) Color
  Transform
• Describes colors in terms that we can more
  readily understand

9

• Hue corresponds to color, saturation corresponds
  to the amount of white in color, and lightness is
  the brightness
• For example a deep, bright orange color would
  have a large intensity (bright), a hue of
  orange , and a high value of saturation
  (deep(??))
• But in terms of RGB components, this color would
  have the values as R 245, G 110, and B20

10
(No Transcript)
11

• Equations for mapping RGB to HSL are
• where

12
Digital Image File Formats

• Bitmap images (raster images) Images that can be
  represented by our image model, I (r,c)

13

• Image file header (????) A set of parameters
  (??) found at the start of the file image and
  contains information regarding
• Number of rows (??)(height, ?)
• Number of columns (??)(width, ?)
• Number of bands (???)
• Number of bits per pixel (?????? ??)(bpp)
• File type (????)

14

• Look-up table (LUT) Used for storing RGB values
  for 8-bit color images

15

• Common image file formats are
• BIN, RAW
• PPM,PBM,PGM
• BMP
• JPEG
• TIFF
• GIF
• RAS
• SGI
• PNG
• PICT, FPX
• EPS
• VIP

16
Matlab ????I/O??????

• ?Matlab?,?????(pixel)?????0?1???????1????,0?????
• ???????RGB,??? red(?) ?green(?) ?blue(?)???????
• ?????????????,??????,?????????

17
Matlab ????I/O??????

• ?????????????,?????????????????RGB ??????????

18
Matlab ????I/O??????

• Show ???imshow( )
• ??????imshow(???? A,?? N) ,?????? A?N???????????
• ?N???,??24?????,???256???
• ????A?????A( , , 3)???????,A( , ,
  1)??????? A( , , 2)??????? A( , ,
  3)????????

19
Matlab ????I/O??????

• ????????????,??????????????,???????????????imshow
  (???? A, lim_l lim_h)
• ?????? A ??????lim_l,????????lim_h,??????

20
Matlab ????I/O??????

• ???????????,????Matlab?workspace?,??imread(????)
  ??,??imshow( )???????
• imwrite( )????????????,??????imwrite(????,
  ????????, ????)

21
Ex2_1.m

• clear close all
• Aimread('1.bmp')
• figure ?????
• imshow(A)
• size(A)
• figure
• imshow(A(,,2)) ?show????
• imwrite(A(,,2),'ex2_1.tif','tif')
• BA(100150,150200,1) ????????
• figure
• imshow(B)
• figure
• imshow(B,100 200)

22
Image Compression using Artificial Neural Networks
23
Image compression using artificial neural networks

• Conventional compression techniques are designed
  for low noise environments, i.e., bit error rate
• Neural networks are suitable for high noise
  environments, i.e., bit error rate

S/N for JPAG algorithm under compression ratio 81
Bit error rate s/n (dB)
0 37.9
0.01 7.1
24
Image compression using artificial neural networks
10?(256x256, 24bits)????
25
Image compression using artificial neural networks
256
8
8
256
3 64-8-64 networks
26
Image compression using artificial neural networks
Original image
Decompressed image
33.0405db
Peak S/N
30.1983db
27
Image compression using artificial neural networks
Comparison of the S/N under compression ratio 81
algorithm s/n (dB)
SBC 23
DCT 21.9
Single-structure NN 23.3
Parallel-structure NN 27.1
New method 30. 5
28
????

• ????????????---??????
• ?????????????????

29
???????????

• ?????

?????????1?????????????,???1?????????????2?
?1 ????
?2 ??????
30
???????????

• ???????,??????????,???????????????
• ????(block)???????????????gray level
  ?????????(0ltthresholdlt1)

31
???????????
Threshold 0.27
S????????? ????block???? ?????????? ???Block?size
32
???????????
Threshold 0.5
Threshold ??, block ????
33
???????????
Threshold 0.75
Threshold ??, block ????
34
?????????

• Conventional compression techniques are designed
  for low noise environments, i.e., bit error rate
• Neural networks are suitable for high noise
  environments, i.e., bit error rate

S/N for JPAG algorithm under compression ratio 81
Bit error rate s/n (dB)
0 37.9
0.01 7.1
35
?????????
256
8
8
256
64-8-64, 16-4-16, and 4-2-4 networks
36
?????????
Block size
64 (8 x 8)
??
16 (4 x 4)
4 (2 x 2)
??
Depth first search
37
?????????
Comparison of the S/N under compression ratio 81
algorithm s/n (dB)
SBC 23
DCT 21.9
Single-structure NN 23.3
Parallel-structure NN 27.1
New method 31. 5