Why the images need to be registered: Differences in camer

1
Multivariate Analysis of Multispectral Images
Derrick Ankomah-Nyarko REU Student Berea
College Dr. Richard Ulrich REU Research
Mentor University of Arkansas 5th Space and
Planetary Sciences REU Research July 27,
2006 Fayetteville
2
Project Synopsis

• Image Definition and Representation
• Coordinate Systems of Images
• Types of Images (relevant to Project)
• Programming Details
• Image Registration
• Stacking and Display of Images
• Display of Spectral Intensity Distribution
• Control FTIR Rock Sample Analysis

3
Image Definition and Pixel Location

• Digital Image

f(x, y)
2-Dimensional array
Height (pixels)
A Pixel
Width (pixels)

• Pixel Coordinate System

• Spatial Coordinate System

Discrete point X(1,1)
Square Patch Y(1.5,1.5)
Image Y
Image X
4
Types of Images

• Color Intensity Image

• Pixel 1 value (varies between shades of
  black(255) and white(0))
• Filter wavelengths range
• 432 nm 1009 nm

• Truecolor Image

• Falsecolor Image

• Pixel 3 values
• (pseudo-red(PR),
• pseudo-green(PG),
• pseudo-blue(PB))
• PR filter 1009 nm
• PG filter 904 nm
• PB filter 436 nm

• Pixel 3 values
• (red(R), green(G),
• blue(B))
• R filter 650 nm
• G filter 510 nm
• B filter 475 nm

5
Image Registration

• Why the images need to be registered
• Differences in camera angle, distance and
  orientation
• Differences in filter wavelengths

6
Types of Spatial Transformation
7
Resizing and Stacking Images

• Image Resizing padarray function

0 0 0 0 0 0 2 5 4 0 0
3 4 3 0 0 4 2 3 0 0
0 0 0 0

• 5 4
• 4 3
• 2 3

padarrary(X, 1 1, 0)
Image X (3 x 3 matrix)
Image X (5 x 5 matrix)

• Stacking Images cat function

cat(3,registered image with shortest wavelength,
,registered image with longest wavelength)
8
Displaying Stacked Image

• Why stacked image is reduced to a tri-layered
  image
• Computer Screen
• Difficulty identifying R, G, B layers for images
  containing more 3 layers
• Human Eye
• The human eye detect colors between wavelengths
  of 380 nm and 780 nm

Pseudo-red layer
Pseudo-green layer
Pseudo-blue layer
cat function
Falsecolor image
9
Image Profiles and Pixel Intensity Information

• Graphic User Interface
• Generates profiles of images
• Generates pixel intensity

Image Profile
Pixel Intensities
10
FTIR Rock Sample Analysis
Known rock sample
FTIR
FTIR Spectrogram
11
Conclusion
12
References

• Gonzalez, Rafael et al. Digital Image
  Processing Using MATLAB. Prentice Hall, 2003.
• Image Processing Toolbox. http//www.mathworks.c
  om/access/helpdesk/help/toolbox/images/
• Gilat, Amos. MATLAB An Introduction With
  Applications. John Wiley Sons, Inc., 2005.
• Hanselman, Duane and Littlefield, Bruce.
  Mastering MATLAB 7. Pearson Prentice Hall, 2005.