Digital Images

GSP 216
Spring 2015
Spatial Data Models
There are two basic ways to represent spatial data:
• Raster
• Raster data represent the feature as a rectangular
matrix of square cells.

• Vector
• Vector data represent features as discrete points,
lines, and polygons.
Vector Data
• Vector data comes in the form of points and lines
that are geometrically and mathematically
associated.
• Points are stored using the coordinates, for example,
a two-dimensional point is stored as (x, y), lines are
stored as a series of point pairs, where each pair
represents a straight line segment.
• Smaller file size compared to rasters
• Example: Shapefiles
Raster Data
• Matrix of cells (or pixels) organized into rows and
columns (or a grid) where each cell contains a value
representing data
• Raster data is used to represent:
• Thematic data (also known as discrete), features such
as land-use or soils data.
• Continuous data, representing phenomena such as
temperature, elevation or
• Spectral data such as satellite images and aerial
photographs
Remotely Sensed Data
Most remotely sensed data we look at represents the
amount of light reflected off the ground and reaching
the satellite sensor.
• Often cover large areas
• Multiple images (bands) are collected simultaneously
for each place in an image

Therefore most of the remote sensed data we will be

working with will be raster data sets
Raster Data
We know that raster data is a matrix of cells (or pixels)
organized into rows and columns (or a grid) where each
cell contains a value representing data
Cell values can be:
• Integers - best used to represent categorical
(discrete) data
or
• Floating‐point values to represent continuous
surfaces
Digital Numbers
• Cell values are also called a digital number or
DN.
• In most of the imagery we work with the DN
represents the strength of the signal (amount of
light) that is assigned to each grid cell (pixel).

• Low or None - Lowest DN (0 is at bottom of scale)

• High - Maximum value (depends on radiometric
resolution)
• Others - Scaled in between (number of possible
increments depends on radiometric resolution)
Digital Numbers
41 71 105

46 29 40

50 41 31
Image Bands
• You can think of image bands (also called channels or
layers) as a collection of pictures taken simultaneously of
the same place
• each of which measures reflected light from a different
part of the spectrum.
Image Bands
For example the Landsat 8 data we looked at in Lab 3 had a
total of eight reflective bands

Band (Layer) Wavelength (nm) Common Name

Coastal aerosol (blue
1 430-450 visible)
2 450-510 Blue visible
3 530-590 Green visible
4 640-670 Red visible
5 850-880 Near infrared (NIR)
6 1570-1650 SWIR1 (Mid-IR)
7 2110-2290 SWIR2 (Mid-IR)
Assigning Bands to Primary
Colors
• Computer monitor uses red, green, and
blue to create color images
• You assign your choice of satellite band
to each primary color
• Brightness of each color is determined
by each pixel value in each band
• Result is a color image with each pixel’s
color determined by combination of
RGB of different brightness.
Primary Colors of Light - Mixing
of Colors
• https://www.youtube.com/watch?v=ksS-
FJ3B8Og

RGB Monitor
Image Bands
• Single Band – Black and White Photography
(panchromatic)
• In this image pixel value
range from 0 to 255

• Higher pixel value (or DN)

is lighter, represents higher
reflectance

• Lower pixel value (or DN)

is darker, represents lower
reflectance
Image Bands
• Multiple Bands - Multispectral images
True Color Composite:
We have the visible three
bands
• Red Band – Red
Channel
• Green Band – Green
Channel
• Blue Band – Blue
Channel

Together they form a

color composite
True Color Composite
Image Bands
False Color Composite:
Any other band combination other than
true color: NIR color composite
• Near Infrared Band – Red Channel
• Red Band– Green Channel
• Green Band – Blue Channel

Together they form a false color composite

False Color Composite
False Color Composite
Shortwave Infrared Composite (6,5,2 or 7,5,2)
• A shortwave infrared composite image is one that contains at
least one shortwave infrared (SWIR) band. Reflectance in the
SWIR region is due primarily to moisture content.

• SWIR bands are especially suited for camouflage detection,

change detection, disturbed soils, soil type, and vegetation
stress.
• Used to show floods or newly burned land.
False Color Composite
False Color Composite
• Shortwave infrared, near infrared, and red composite
• Shortwave infrared Band –Red Channel
• Near infrared Band – Green Channel
• Red band- Blue Channel
• Distinguish ice from snow, water, and clouds
• Ice is pale blue (thicker ice is brighter), open water is
navy, snow is blue-green, and clouds are white or blue-
green (depending on temperature and composition).
False Color Composite

Image: http://earthobservatory.nasa.gov/ MODIS February 19, 2014

False Color Composite

• Shortwave infrared
Band –Displayed in Red
• Near infrared Band –
Displayed in Green
• Red band- Displayed in
Blue
• MODIS Bands 7,2,1
Red Band NIR Band SWIR Band
(displayed (displayed (displayed
in Blue) in Green) in Red)
Landsat 8 Bands
Band
Band 1 – coastal aerosol
Band 2 – blue
Band 3 - green
Band 4 - red
Band 5 - Near Infrared (NIR)
Band 6 - Short-wave Infrared
Band 7 - Short-wave Infrared
Band 8 - Panchromatic
Band 9 – Cirrus
Band 10 – TIRS 1
Band 11 – TIRS 2
Wavelength Useful for mapping
0.43-0.45
Coastal and aerosol studies
Bathymetric mapping, distinguishing soil from
0.45-0.51 vegetation and deciduous from coniferous
vegetation
Emphasizes peak vegetation, which is useful for
0.53-0.59
assessing plant vigor
0.64-0.67 Discriminates vegetation slopes
085.-0.88 Emphasizes biomass content and shorelines
Discriminates moisture content of soil and
1.57-1.65
vegetation; penetrates thin clouds
Improved moisture content of soil and vegetation
2.11-2.29
and thin cloud penetration
.50-.68 15 meter resolution, sharper image definition
1.36 -1.38 Improved detection of cirrus cloud contamination
100 meter resolution, thermal mapping and
10.60 – 11.19
estimated soil moisture
100 meter resolution, Improved thermal mapping
11.5-12.51
and estimated soil moisture
Histograms
• A histogram is a graph showing the number of
pixels in a single band corresponding to each
possible DN.

• Histograms used for image enhancements

including: contrast stretching, atmospheric
correction, statistical manipulations
Histograms
Resolution
Resolution is characterized in several different
ways:

• Spatial
• Spectral
• Radiometric
• Temporal
Spatial Resolution
• Usually reported as the length of one side of a
single pixel
• In analog imagery (film), the dimension (or
width) of the smallest object on the ground that
can be distinguished in the imagery
• Determined by sensor characteristics (for digital
imagery), film characteristics (for air photos)
including field of view, altitude, etc.
Spatial Resolution
• Coarse or low resolution - images where only
large features are visible

• In fine or high resolution images, small objects

can be detected
Spatial Resolution
low resolution high (fine) resolution
Satellite Data Spatial Resolution

• MODIS: 250 - 1000 m

• Landsat MSS: 80 m
• Landsat 5, 7, 8: 30 m (15 m panchromatic)
• SPOT: 20 m
• ASTER: 15m
• Digital Globe 0.4 m (0.3m next year)
Spatial Resolution

Landsat 8: 30 m
30 m

30 m

Each Landsat 8
pixel is 30m x
30m or 900m2
Spatial Resolution
WorldView 3 (DigitalGlobe):
0.4 (0.3) m 0.4 m

0.4 m

Each WorldView-3
pixel is 0.4m x 0.4m
or 0.16m2
Spectral Resolution
• How many spectral “bands” an instrument
records
• How “wide” each band is or the range of
wavelengths covered by a single band
• Landsat 8 has 11 bands, including several in the
infra-red spectrum, the bands range from a
spectral resolution of 0.07 to 2.1 μm
Landsat Spectral Resolution
MODIS Spectral Resolution

Moderate Resolution Imaging Spectroradiometer

(MODIS)
Radiometric Resolution
• Radiometric resolution is how finely does the satellite
(or sensor) divides up the radiance it receives in each
band

• The range of different intensities of radiation the sensor

is able to distinguish.

• Usually expressed as number of bits used to store the

maximum possible DN value
• 8 bits = 28 = 256 levels (usually 0 to 255)
• 16 bits = 216 = 65,536 levels (0 to 65,535)
Radiometric resolution
• The intensity value detected by the sensor needs
to be scaled to fit within this range of value.

• In a Radiometrically Calibrated image, the actual

intensity value can be derived from the pixel
digital number.
Radiometric Resolution

8 bits = 28 = 256
levels (usually 0 to
255)

2 bits = 22 = 4 levels
(usually 0 to 3)
Radiometric Resolution

1 bit = 21 = 2 levels
(0 or 1, black or
white)
Radiometric resolution
• 1 bit ( 0 - 1)
• 8 bit ( 0 - 255 ) (older Landsats, many others)
• 16 bit ( 0 - 65,535 ) (Landsat 8)

8 bit data (256 values)

• Everything will be scaled from 0 – 255
• Subtle details may not be represented

16 bit data (65,536 values)

• Wide range of choices
• Required storage space will be twice that of 8 bit
Color Display
• Each color (red, green, blue) is 8 bits, for a total of 24 bits
• Since each color has 256 shades, we can multiply 256 for
red, times 256 for green, times 256 for blue and get:
256 x 256 x 256 = 16,777,216 colors
• The human eye can only discern about 10 million
different colors, so 24 bit is known as “True Color”
 Temporal resolution
• Time between two subsequent data
acquisitions for an area or “return time”

•All of the Landsat satellites have a 16-day

return time

•MODIS has a 1-2 day return time.

Return Time
(Temporal Resolution)
• Depends on:
• Orbital characteristics
• Swath width
• Ability to point the sensor
Orbital Characteristics

• Geosynchronous
• Polar/ Sun
synchronous
Geosynchronous Orbits
• Satellite orbits the earth at a rate that allows it to
match the earth’s rotation— the satellite is always
over the same place
• Narrow range of altitudes—about 35,786 km above
the equator.
• Useful for communications, weather etc.
• Example: GOES satellite (weather)
• Geosynchronous orbiting earth satellite
Polar/Sun Synchronous Orbits
• Pass roughly over the north and south poles
• Fly over the same place on earth at the same time
of day
• Examples: Landsat, AVHRR
• Good for land remote sensing
• Return time related to spatial resolution, latitude,
swath width, and orbital altitude
Landsat Orbit

http://www.youtube.com/watch?v=P-
lbujsVa2M
Metadata
Spectral Reflectance
Spectral Reflectance
Spectral Reflectance