Image Acquisition (Ch. 3)

Question 1

Spatial Resolution (Four Main Resolutions in RS)

Answer 1

Determines the smallest object that can be detected in an image.

• typically expressed as pixel size 1 x 1 meters
• in general, can only identify features that are larger than the pixel size
• higher spatial resolution = smaller pixel size
• Smaller pixel size usually means more pixel homogeneity or pure pixel (non-mixed pixels? which means more pure spectral signatures.

Question 2

Radiometric Resolution (Four Main Resolutions in RS)

Answer 2

Sensor sensitivity (capacity to discriminate small differences in recorded EM radiation). Determines the number of grey level values that can be distinguished on an image.

• Output from sensor is recorded as numeric values (DN).
• Rad. Res: sensitivity to detect small differences in reflected or emitted energy
• Each brightness level (DN value) corresponds to an amount of recorded EM energy
• Determines the range of DN values in each band (maximum number of brightness levels)

Finer radiometric resolution -> more sensitivity to detect small differences in reflected or emitted energy -> more brightness levels available -> larger image (more storage space & processing time required)

Radiometric resolution is expressed in terms of the
number of bits (binary values) used to code the
energy measured by the sensor:
– Early sensors: 128 values (7 bits)
Landsat: 256 values (8 bits)
– IKONOS & Quickbird: 2048 (11 bits)
– MODIS: 4096 (12 bits

DN range = 2# (# = number of bits)

Question 3

Spectral Resolution (Four Main Resolutions in RS)

Answer 3

Number of bands and their width (wavelength range). Usually more bands and/or narrower is better (all other things being equal).

• Panchromatic: only one band, in greyscale
• Multispectral: several bands, images can be displayed in color (using color composites)
• Hyperspectral: 100s of band.

More and narrower bands is more discrimination power.

Question 4

Temporal Resolution (Four Main Resolutions in RS)

Answer 4

Timeframe of the repeat cycle of the satellite’s return to record the same area of the earth’s surface.

Timeframe of the repeat cycle of the satellite’s return to record the same point on the earth’s surface.
• Monthly, bi-weekly … (e.g., Landsat)

Question 5

Sensors with High, Medium, and Low Spatial Resolution

Answer 5

High Spatial Resolution:
- Quickbird, IKONOS, Geoeye1,

Medium Spatial Resolution:
- Landsat series, 30 m visible

Low Spatial Resolution:
- Terra-Aqua Satellites e.g., MODIS. Lowest resolution

Question 6

Image Composites

Answer 6

• 3 bands are displayed as a single image
• Very useful to facilitate the identification of different earth features
• Different band combinations show the terrain differently

Question 7

Type of Orbit: Polar vs. Geostationary

Answer 7

Geostationary satellites:

• Always observe the same area at high equatorial orbits (36,000 km above Earth)
• Register an entire visible hemisphere in one image
• High temp. resolution, low spatial resolution

Polar-orbiting satellites:

• Cover the entire planet in hours or days at lower near polar orbits.
• Sun-synchronous orbits: observe the same area at the same local sun time (removes illumination effect due to differences in time of day -> multi-temporal comparisons)
• Higher spatial resolution in general