7/6/2018

Preprocessing of RS Imagery
 Preprocessing refers to those operations that
are preliminary to the principal analysis of
remotely sensed data
GeoTech 103: Basic  Preparatory phase of RS data
Remote Sensing  Improves image quality
 Key procedures:
• Radiometric preprocessing: adjust digital values
Preprocessing of Remotely for effects of a hazy atmosphere
Sensed Imagery • Geometric preprocessing: bring an image into
registration with a map or another image
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])

Preprocessing of RS Imagery Preprocessing of RS Imagery

Radiometric Preprocessing Radiometric Preprocessing
 Controls the brightness values of an image
to correct for sensor malfunctions or to
adjust the values to compensate for
atmospheric degradation
• Brightness from Earth’s surface
• Brightness from the effects of atmospheric
scattering
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])
 7/6/2018

Preprocessing of RS Imagery Preprocessing of RS Imagery

Preprocessing for atmospheric correction Preprocessing for atmospheric correction
 Histogram minimum method (HMM) or the dark object
 Radiative transfer code (RTC): computer
subtraction (DOS) technique:
models, which model the physical behavior
of solar radiation as it passes through the
atmosphere
 Image-based atmospheric correction:
examines the spectra of objects of known or
assumed brightness recorded by
multispectral imagery
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])

Preprocessing of RS Imagery Preprocessing of RS Imagery

Calculating Radiances from DNs Estimation of Top of Atmosphere Reflectance
 Digital numbers (DNs): pixel values in integer  Reflectance: proportion of radiation reflected
• Facilitate computation and transmission from objects relative to the amount of the
same wavelengths incident upon the object
• Scale brightnesses for display
 Useful for defining the distinctive spectral
• Express relative brightnesses within an image
characteristics of objects
 DNs do not present brightnesses in the
physical units (Watts per square meter per Observed brightness
micrometer per steradian) Reflectance =
Irradiance
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])
 7/6/2018

Preprocessing of RS Imagery Preprocessing of RS Imagery

 Destriping: application of algorithms to  Feature extraction: identification of specific
adjust incorrect brightness values to values bands or channels that are of greatest value
thought to be near the correct values for analysis
• Features: statistical characteristics of image data
• Features: not geographical features visible on an
Sixth-line striping in image
Landsat MSS data  Feature selection: selection of image
information that is tailored for a particular
application
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])

Preprocessing of RS Imagery Preprocessing of RS Imagery

Feature selection by principal components analysis Geometric Correction by Resampling
 Resampling: scales, rotates, translates, and
performs related manipulations as necessary
to bring the geometry of an image to match a
particular reference image of desired
properties
 Image resampling: use interpolation to bring
an image into registration with another image
or a planimetrically correct map
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])
 7/6/2018

Preprocessing of RS Imagery Preprocessing of RS Imagery

Geometric Correction by Resampling Geometric Correction by Resampling

Open circles (o) – reference

Nearest-neighbor resampling
grid of known values in the
input image Each estimated value (•)
receives its value from the
Black dots (•) – regular grid
nearest point on the
of points to be estimated to
form the output image reference grid (o)

GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])

Preprocessing of RS Imagery Preprocessing of RS Imagery

Geometric Correction by Resampling Ground Control Points (GCPs)
 Georeferencing is the matching of an image to
reference points that correspond to known  GCPs are features that can be located with
locations on the ground precision and accuracy on accurate maps
 Georeferencing includes matching of an image to yet that are also easily located on digital
a reference image images
 Georeferenced images:
• Specific geographic projection with defined projection • Small as a single pixel
and coordinate system
• Spectrally distinct areas as small as a few
• Important for locational referencing
pixels
• Important for referencing other images
GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])
 7/6/2018

Preprocessing of RS Imagery Preprocessing of RS Imagery

Ground Control Points (GCPs) Data fusion
 Intersections of major  Edges of land-cover  Data fusion refers to processes that bring
highways parcels
images of varied resolution in to a single
 Distinctive water bodies  Edges of stream
junctions
image
• Incorporates the high spatial resolution of a
panchromatic image with the multispectral
content of a multiband image at coarser
resolution

GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])

Preprocessing of RS Imagery Preprocessing of RS Imagery

Image Data Processing Standards Level Characteristics
Level Characteristics Derived geophysical variables (e.g., ocean wave height, soil
Reconstructed, unprocessed instrument and payload data at full 2 moisture, ice concentration) at the same resolution and
0 resolution, with any and all communications artifacts (synchronization location as Level 1 source data
frames, communications headers, duplicate data) removed

Reconstructed, unprocessed instrument data at full resolution, time Variables mapped on uniform space–time grid scales, usually
referenced, and annotated with ancillary information, including with some completeness and consistency (e.g., missing points
radiometric and geometric calibration coefficients and georeferencing 3
1a interpolated, complete regions mosaicked together from
parameters computed and appended but not applied to the Level 0 data
(or if applied, in a manner that Level 0 is fully recoverable from Level 1a
multiple orbits)
data)

Level 1a data that have been processed to sensor units (e.g., radar Model output or results from analyses of lower-level data
1b backscatter cross section, brightness temperature); not all instruments 4 (i.e., variables that were not measured by the instruments
have Level 1b data; Level 0 data is not recoverable from Level 1b data but instead are derived from these measurements)
GeoTech103: Basic Remote Sensing
Source: Parkinson et al. (2006, p. 31) Prepared by: Gift Dumedah ([email protected]) GeoTech103: Basic Remote Sensing
Source: Parkinson et al. (2006, p. 31) Prepared by: Gift Dumedah ([email protected])
 7/6/2018

QUESTIONS!

GeoTech103: Basic Remote Sensing Prepared by: Gift Dumedah ([email protected])