Digital Image Processing

							Digital Image Processing

            Ligang Liu
        Zhejiang University
      ligangliu@zju.edu.cn
Media
A picture is worth 1000 words…
A video is worth 1000 sentences…
     Rich information from visual data
     Examples of images around us
         Natural photographic images
         Artistic and engineering drawings
         Scientific images (satellite, medical, etc.)
     Motion picture –video
         Movies, TV programs, news
         Family video
         Surveillance and highway camera
    Why do we process images?
   Enhancement and restoration
       remove artifacts and scratches from an old
        photo/movie
       improve contrast and correct blurred images
   Transmission and storage
       images from oversea via Internet, or from a remote
        planet
   Information analysis and automated recognition
       providing “human vision” to machines
   Security and rights protection
       encryption and watermarking
    Why Digital?
   “Exactness”
       Perfect reproduction without degradation
       Perfect duplication of processing result
   Convenient & powerful computer-aided processing
       Can perform rather sophisticated processing through hardware
        or software
       Even kindergartners can do it!
   Easy storage and transmission
       1 CD can store hundreds of family photos!
       Paperless transmission of high quality photos through network
        within seconds
Human Vision System
   Image is to be seen.
   Perceptual Based Image Processing
       Focus on perceptually significant information
       Discard perceptually insignificant information
   Issues:
       Biological
       Psychophysical
Color
   Color is the perceptual result of light
    having wavelength 400 nm to 700 nm
    that is incident upon the retina.
   “Power distribution exists in the physical
    world, but color exists only in the eye
    and the brain.”
   Does “red” mean the same to different
    people?
Color Spectrum
Grassman's First Law of
Additive Color Mixture
   Any color can be matched by a linear
    combination of three other colors
    (primaries, eg RGB), provided that none
    of those three can be matched by a
    combination of the other two.
       C= Rc(R ) + Gc(G) + Bc(B)
Color Spaces
   RGB
   CMY
   CIE XYZ
   s 
    l
Different Image Types
   Binary images (0 or 1)
   Gray images (0~255)
   Color images
       indexed color images
       full color images (24 bits per pixel, 8-red,
        8-green, 8-blue) )
A Binary Image
Gray Images
   8 bits per pixel
Full Color Images
   24 bits per pixel, and the three
    channels R G B are three gray images
    respectively
Color Components
Image Programming
   class CImage
   {
        unsigned int width;
        unsigned int height;
        unsigned char *data;
   };

   Not difficult…
Related Fields
   Imaging
       Medical, remote sensing, weather
   Computer vision
   Computer graphics
   Machine learning
   Video processing
Related Math
   Fourier analysis
   Wavelet
   Probability and statistics
   PDE
   Linear/nonlinear optimization
   Machine learning
   …
    References
   Journals
       IEEE Transaction on Pattern Analysis and Machine
        Intelligence (PAMI)
       IEEE Transaction on Image Processing
       IEEE Transaction on Signal Processing
       IEEE Transaction on Circuits and System for Video
        Technology
       International Journal on Computer Vision (IJCV)
       Pattern Recognition
   Conferences
       Graphics Conferences (Siggraph…)
       Vision conferences: ICCV, ECCV, ACCV, CVPR
Course Information
   Seminar
       Report papers by yourselves
   Grading
       Seminar reports and final report
   Course homepage and FTP
Objectives
   Learn something interesting
   Do something interesting
   Find some interesting problems

   Improve your abilities and experiences!
Requirements
   Reporter
       Over-prepared: read a series of important papers, PPT (texts
        and images)
       Professional: PPT, explaining, interaction…
       List all references on the last slide
       His own idea or own work
   Audience
       Challenging the reporter
       Ask questions
       Learn something new
   Active and creative!
Q&A