27
Sensor Resolution

Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Sensor Resolution

Page 2: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Outline for 4/14/2003

• Digital imagery - follow-on from last lecture

• Spatial Resolution

• Spectral Resolution

• Temporal

• Radiometric

• Instrument sensitivity

Page 3: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Digital Number

imaging opticsdetectors

electronics

at-sensorradiance DN

DN is proportional to at-sensor radiance

Page 4: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Resolution and InstrumentResponse Functions

• Sensor has finite precision

• Input signals vary in time and space

• Sensor has “response function” (spatial,spectral)

inputsignal

outputsignal

convolutionw/responsefunction

Page 5: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Spatial Resolution• “A measure of the smallest angular or linear

separation between two objects that can beresolved by the sensor”. (Jensen, 2000)

• Resolving power is the ability to perceive twoadjacent objects as being distinct– size– distance– shape– color– contrast characteristics– sensor characteristics

Page 6: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

• Instantaneous field of view (IFOV) is theangular field of view of the sensor,independent of height

• IFOV is a relative measure because it is anangle, not a length

b

Page 7: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

GIFOV

• Ground-projected instantaneous field of view(GIFOV) depends on satellite height (H)

GIFOV = 2H tan IFOV2

Ê

Ë Á

ˆ

¯ ˜

Page 8: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

1 meter resolution 250 meter resolution

IKONOS image of Gunnison River Basin, CO1

kilo

met

er

Page 9: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number
Page 10: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Spectral Resolution

• The width and number of spectral intervals inthe electromagnetic spectrum to which aremote sensing instrument is sensitive

• Allows characterization based on geophysicalparameters (chemistry, mineralogy,etc.)

Page 11: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Spectral Resolution

• Determined by:– the number of spectral bands

– spectral response function of each band

– full-width at half-maximum (FWHM)

Page 12: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

AVIRIS image of Moffat Field, CA

224 channels from 0.4 - 2.5 mm10 nm bandwidth

Page 13: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

• Surface components with very distinctspectral differences can be resolved usingbroad wavelength ranges

Page 14: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

vegetation spectral signatures from Jasper Ridge

Subtle differences require finer spectral resolution

Page 15: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number
Page 16: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Radiometric Resolution

• Number of digital levels that a sensor can useto express variability of brightness within thedata

• Determines the information content of theimage

• The more levels, the more detail can beexpressed

Page 17: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Radiometric Resolution

• Determined by the number of bits ofwithin which the digital information isencoded

22 = 4 levels28 = 256 levels212 = 4096 levels

Page 18: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

2 b

it r

adio

met

ric

reso

luti

on

8 b

it r

adio

met

ric

reso

luti

on

Page 19: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Imag

e B

right

ness

Scene Brightness

DynamicRange

ActualSensorResponse

IdealResponse

DarkCurrentSignal

Saturation

Page 20: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Temporal Resolution

• The frequency of data acquisition overan area

• Depend on:– the orbital parameters of the satellite

– latitude of the target

– swath width of the sensor

– pointing ability of the sensor

Page 21: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

• Multi-temporal imagery is important for– infrequent observational opportunities (e.g.,

when clouds often obscure the surface)

– short-lived phenomenon (floods, oil spills,etc.)

– rapid-response (fires, hurricanes)

– detecting changing properties of a feature todistinguish it from otherwise similar features

Page 22: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Breakup of the Larsen B Ice Shelf

MODISimagery fromJanuary 31, 2002-March 6, 2002

Courtesy of Ted Scambos, NSIDC

Page 23: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Signal Strength

• Depends on– Energy flux from the surface

– Altitude of the sensor

– Spectral bandwidth of the detector

– IFOV

– Dwell time

Page 24: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Signal-to-Noise Ratio (SNR)Sensor responds to a both target brightness

(signal) and electronic errors from varioussensor components (noise)

SNR = signal to noise ratio

signal = the actual energy reaching the detectornoise = random error in the measurement (all

systematic noise has been removed)

To be effective, sensor must have high SNR†

signalnoise

Page 25: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Noise = iDN - mDN( )2

n -1i=1

n

Â

Page 26: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

200201199203202201200

50%

Mean DN = 201Noise = 1.345SNR = 201/1.345 = 149

Page 27: Sensor Resolution · •Digital imagery - follow-on from last lecture •Spatial Resolution •Spectral Resolution •Temporal •Radiometric •Instrument sensitivity. Digital Number

Noise Equivalent Radiance orReflectance

• A measure of the lowest signal that can bedetected just before the signal falls below thelevel of the noise

NEDL or NEDr = the standard deviationof the Mean (of a set of measurements)that produces a SNR of 1