1. Initial Display Alternatives and Scientific Visualization
Geography KHU Jinmu Choi
Scientific Visualization
B-W Hard Copy Image Display
Temporary Video Image Display
Optimum Band Combination
Merging RS Data
Distance & Area Measurement
Next…
Remote Sensing

2. 1. Initial Display Alternatives and Scientific Visualization
“visually exploring data and information to gain understanding and insight into the data”
Presentation graphics:
Primarily concerned with the communication of information and results that are already understood
Seeking to understand the data and gain insight
Remote Sensing

3. Scientific Visualization
Remote Sensing
(source: Jensen, 2011)

4. Input and Output Relationships
Brightness map: a computer graphic display of the brightness values, BVi,j,k,
Ideally, one-to-one relationship between input brightness values and the resultant intensities on the display
Remote Sensing
(source: Jensen, 2011)

5. 2. Black-and-White Hard-copy Image Display
Line Printer/Plotter Brightness maps
Laser or Ink-jet Printer Brightness maps
Remote Sensing

6. Class Intervals with Line Printer Brightness Maps
(source: Jensen, 2011)
Natural Breaks
Remote Sensing
Equal Size
Equal Area

7. Symbolization and Perceived Grayness of Line Printer
(%) .
34.5 O=
54.4 X
44.8 MW
55.9 X-
48.5
TVA
57.6 Z=
50.5
HIXO
64.1
Remote Sensing

8. 3. Temporary Video Image Display
Bitmapped Graphics
RGB Color Coordinate System
Color Look-up Tables: 8-bit
Color Look-up Tables: 24-bit
Color Composites
Remote Sensing

9. Temporary Video Image Display
Bitmapped Graphics: a pixel brightness value at each row and column in a matrix as being bitmapped image
Remote Sensing
(source: Jensen, 2011)

10. RGB Color Coordinate System
A Red-Green-Blue (RGB) color coordinate system based on additive color theory (light is mixed)
Printer: Subtractive color theory (pigments are mixed)
Using three 8-bit images and additive color theory, we can display 224 = 16,777,216 color combinations
Remote Sensing

11. Color Look-up Tables: 8-bit
Gray tone or color of an individual pixel on a screen
Controlled by the size and characteristics of a separate bank of computer memory called a color look-up table
Color look-up table
Exact disposition of each combination of red, green, and blue values associated with each 8-bit pixel
With color look-up tables how black-and-white and color density-sliced brightness maps are produced
Remote Sensing

12. 24-bit Digital Image Processing System
Color look-up table :
exact disposition of each combination of red, green, and blue values associated with each 8-bit pixel
(source: Jensen, 2011)
Remote Sensing

13. Color Density Slicing, Visible
Color Class
Interval
Visual
Color
Color Lookup Table Values Red, Green, Blue
Brightness Value
Low High 1
Cyan
0, 255, 255
Shade of gray
17, 17, 17
18, 18, 18
19, 19, 19 *
59, 59, 59 2
Red
255, 0, 0
Remote Sensing

14. Color Density Slicing, Thermal
Color Class
Interval
Visual
Color
Color Lookup Table Values
Red, Green, Blue
Apparent Temperature
Low High
Brightness Value Low High
1. Land
gray
127,
2. River Ambient
Dark blue
0, 0, 120 C
Light blue
0, 0, 255
– 2.8 C
Green
0, 255, 0
– 5.0 C
Yellow
255, 255, 0
– 10.0 C
Orange
255, 50, 0
– 20 C
Red
255, 0 , 0
8. > 20 C
White
255, 255,255
Remote Sensing

15. 24-bit Digital Image Processing System
(source: Jensen, 2011)
To display up to three bands of RS data, use three separate bans of image processor memory and three color look-up tables
Remote Sensing

16. 4. Optimum Index Factor
Ranks the 20 three-band combinations from six bands of Landsat TM data (not TIR band)
The most useful combination: the largest OIF has the most information (by variance) with the least amount of duplication (by correlation)
Band combination:
1,2,3
1,2,4
1,2,5
1,2,6
2,3,4 etc.
Where sk is the standard deviation for band k, and rj is the absolute value of the correlation coefficient between any two of the three bands being evaluated (e.g., 1 vs. 2, 1 vs. 3, 2 vs. 3)
Remote Sensing

17. Optimum Index Factor Landsat TM : 1, 4, 5 combination is the best
(source: Jensen, 2011)
Band combination:
3,4,5
Band combination:
1,2,3
Better because the bands are not highly correlated
Landsat TM : 1, 4, 5 combination is the best
Remote Sensing

18. Sheffield Index The Sheffield Index (SI) is:
(source: Jensen, 2011)
Sheffield Index
The Sheffield Index (SI) is:
where is the determinant of the covariance matrix of subset size p
p = 3 since to discover the optimum three-band combination for image display purposes
Choose the highest determinant of the covariance matrix.
Band combination:
1,2,3
1,2,4
1,2,5
1,2,6
2,3,4
2,3,5
2,3,6
3,4,5
3,4,6 etc.
Remote Sensing

19. 5. Merging Remotely Sensed Data
Band Substitution
Color Space Transformation and Substitution
RGB to IHS Transformation and back again
Chromaticity Color Coordinate System and the Brovey Transformation
Principal Component Substitution
Pixel-by-pixel Addition of High-Frequency Information
Smoothing Filter-based Intensity Modulation Image Fusion
Remote Sensing

20. Merging Different Types of RS Data
Accurately registered to one another
Resampled to the same pixel size
Several alternatives exist for merging the data sets, including:
Simple band substitution methods
Color space transformation and substitution methods using various color coordinate systems
Substitution of the high spatial resolution data for principal component #1
Remote Sensing

21. Intensity-Hue-Saturation (HIS)
Vertical axis: intensity (I): black (0) to white (255)
Circumference: hue (H): wavelength of color.
Saturation (S): purity of the color: 0 at the center of the color sphere to 255 at the circumference
Remote Sensing

22. Intensity-Hue-Saturation (HIS) Substitution
RGB → IHS → I to High Resolution → RGB
IHS ↔ RGB transformation equations:
Substitute Intensity data from the IHS transformation for one of the bands, e.g., RGB = 4, I, 2
Remote Sensing

23. Merging using Brovey Transform
To fuse images with different spatial and spectral characteristics
where R, G, and B are the spectral band images, P is a co-registered band of higher spatial resolution data , and I = intensity
Remote Sensing

24. Principal Component Substitution
Stretched panchromatic image is substituted for the first principal component image because the first principal component image
Chavez et al. (1991)
Principal components analysis to six Landsat TM bands.
The SPOT panchromatic data were contrast stretched.
The stretched panchromatic data substituted for the first principal component image
The data were transformed back into RGB space
Remote Sensing

25. 6. Distance and Area Measurement
(source: Jensen, 2011) c a b
Remote Sensing

26. Distance (Length) of a Linear Feature
This logic may be used to identify the length of the longest axis of the mangrove island, where:
Remote Sensing

27. Area Measurement
Simpson’s rule
(source: Jensen, 2011)

28. Summary Scientific Visualization B-W Hard Copy Image Display
Temporary Video Image Display
Optimum Band Combination
Merging RS Data
Distance & Area Measurement
Remote Sensing

29. Next Exercise: Image Annotation & Map Composition
Lecture: Radiation Principles and Radiometric Correction
Source:
Jensen and Jensen, 2011, Introductory Digital Image Processing, 4th ed, Prentice Hall.