Lab 4: Miscellaneous Image Functions
Goal and Background:
The goals of this lab
were to (1) delineate a study area from a larger satellite image scene, (2)
demonstrate how spatial resolution of images can be increased, (3) introduce
some radiometric enhancement techniques in optical images, (4) link a satellite
image to Google Earth which is used as a source of additional material, (5)
dive into a variety of resampling methods used on satellite images, (6) explore
image mosaicking, and (7) look at binary change detection through graphical
modeling(Wilson, 2020).
Methods and Analysis:
The first
part of the lab had to do with image subsetting and the creation of an Area of
Interest or AOI. A subset in this instance is a small section that is taken
from a larger image creating an AOI polygon. To do this you take a shapefile and
lay it over the raster data and then create a raster image that matches the
shapefile. This is shown in Figure 1.
a
b
Figure 1: Shows the delineation of an AOI from a
larger satellite image. A. Shows the original raster data with the shapefile
layer derived. B. Shows the resulting raster AOI that was cut under Subset and
Chip.
The
second part of the lab was Image Fusion. In this section I created a higher
spatial resolution image from a coarse resolution image by pan-sharpening. The
increase of spatial resolution was done to enhance the visual interpretation later.
The increased spatial image is pixelated because resolution was increased in
the reflective band.
The
third part of the lab involved radiometric enhancement techniques which were
used to enhance image spectral and radiometric quality. I used a haze reduction
analysis which created a darker, clearer image because the haze was reduced.
The
fourth part of the lab we linked our image viewer to google earth. To do this,
google earth is opened and then the image in ERDAS is linked and connected to
the google earth displayer. This is a useful tool for selective image
interpretation because it can show the same AOI but with a better resolution.
Additionally, Google Earth shows the names of buildings and other features
which can be used to help understand the setting of ones AOI.
The
fifth part of the lab had to do with different resampling techniques.
Resampling changes the pixel size. We used a nearest neighbor analysis and
bilinear interpolation with the goal of eliminating stairstepping.
|
|
Figure 2: A shows the original two images to be
mosaicked. B shows the two images after they were mosaiced with Mosaic Express
with a distinct boundary between the two images.
Figure 3: Image after going through Mosaic Pro with less distinct boundaries and a better image than those above.
The
final part of the lab involved Binary Change detection. To do this, I estimated
and mapped the brightness values of pixels that changed in Eau Claire County
and four other neighboring counties between August 1991 and August 2011. The equation below shows the basis for our
model.
ImageΔ = change image.
b = specific band.
Image2 = Brightness values of 2011 image.
Image1 = Brightness values of 1991 image.
C =
constant: 127.5 for an 8-bit image.
Results and Knowledge Gained:
By the end
of this lab I gained skills in image pre-processing, spatial resolution
enhancement, study area delineation, image mosaicking, and graphical modeling
that is all used for remote sensing analytics.Sources:
Data for this lab exercise is in the class folder in
Q/StudentCoursework/Wilson/GEOG338-001/SHARE/Lab 4. Data sources are as
follows: Satellite images are from Earth Resources Observation and Science
Center, United States Geological Survey. Shapefile is from Mastering ArcGIS 6th
edition Dataset by Maribeth Price, McGraw Hill. 2014
Wilson, C. (2020) Miscellaneous image functions, LAB 4,
GEOG 338: Remote sensing of the Environment (pp. 1-35).
Comments
Post a Comment