Final Project

Photo Interpretation and Remote Sensing - Final

The final project for this class was to use what was learned in all the previous classes to answer a question.  I chose to go with the suggested project for which the data was provided.  The question being, Is urbanization to blame for the decrease in clarity of the waters of Lake Tahoe.  

To answer this question a composite image was created of individual bands of a 2010 Landsat 5 TM image in ArcMap, then imported into Erdas Imagine where spectral signatures were created and a supervised classification was run.  This image was then returned to ArcMap to compare to a 1992 National Land Cover Database image.  The comparison showed urbanization is likely the cause of the problem.

It's a relief to finally finish this course as it was the most difficult of them all.  I think part of the problem was I kept looking for things to be black and white, and quite frequently they were more best guess.  It was a difficult concept for me to adjust to.  In the end I think I did okay, and somewhere along the way I also came to enjoy it.  I think that was at the very end, when I knew it was almost over and I'd never likely need to do this again.  But at least now I can look back at this with fond memories.  

Supervised Classification

Photo Interpretation and Remote Sensing - Mod 10

This week we created a land use map of Germantown, Maryland using Supervised Classification tools in Erdas.  Training sites were created using a combination of polygons and the Growing Properties tool.  With these tools, features were selected and defined so an algorithm would know what class to assign to each pixel.  We started by defining 14 classes, then merged them down to 8 using the Recode tool.    Once that was done a Class Name column was added and the classes renamed, then an area column was added to the attribute table to calculate the area covered by each class.  To finish the map off it was added to ArcMap and symbolized appropriately.  

Unsupervised Classification

Photo Interpretation & Remote Sensing - Mod 9

This week's less in Unsupervised Classification was pretty straightforward and simple, though somewhat time consuming.  Most of the work was done in Erdas Imagine starting with running the Unsupervised Classification tool on a high resolution aerial photograph of the UWF campus.  This

resulted in a thematic raster that allowed us to simplify the image into fewer classes by selecting the pixels and changing them.  Our task was originally to classify the image into four categories; Trees, Grass, Building/Roads, and Shadows.  This seemed simple enough until some of the clusters affected multiple features.  In order to deal with this problem a Mixed class was added as well and anything that affected more than one feature could be added to that.  Sometimes the affect on a second feature was so limited it made more sense to keep that class as the first feature.  Once we had the image classified we merged the classes so we had only the five specified.

Next we added fields to the attribute table for Class Name and Area.  This allowed us to calculate the area for each of the classes so we could determine the percentage of permeable verses impermeable surfaces.  Permeable surfaces include Grass and Trees, impermeable, Buildings\Roads.  The other two classes included a combination of both, so before we could make that calculation we first had to calculate what percentage of each of those classes were made of each surface.

Thermal & Multispectral Analysis

Photo Interpretation & Remote Sensing - Mod 8

This week's assignment was to use image manipulation and interpretation techniques to identify a feature from an image using the thermal infrared band as part of the analysis.  The selection was made in Erdas Imagine using the TM Thermal Infrared Composite band combination of red for the thermal layer Band 6, green for Band 3 and blue for Band 2.  In ArcMap the same image was symbolized with a composite combination of red for Band 1, green for Band 2 and blue for Band 3 for a more real world image, which made the feature clearer and more easily identified as an airport.  The coordinate for this feature are 30° 28' 34.8306' N, 86° 31' 7.3224' W and allowed it to be identified as Destin - Ft. Walton Beach Airport (VPS) in northwest Florida.

Image Preprocessing 2: Spectral Enhancement and Band Indices

Photo Interpretation and Remote Sensing - Mod 7

This week was very challenging.  We worked with histograms a lot this week learning how to analyze them in order to interpret images.  Our final exercise was to identify certain areas on the map based on histogram information.  That was very challenging, but by the end I had a little better understanding of this week's lesson.  Here are the maps created:

Image Processing 1: Spatial Enhancement and Radiometric Correction

Photo Interpretation and Remote Sensing - Mod 6

This week we used radiometric and spatial enhancements to enhance an image and reduce striping.  

The first step was to perform a Fourier transformation in order to run some of the Fourier tools in a Fourier Transform Editor.  Prior to running this step the image was just a big white blur with a few black splotches until it was zoomed to 1:150000.  After running the Fourier Transform Editor tools it was a complete image that could be zoomed to its extent, but it still had striping.  I tested numerous tools trying to find that right combination that would lessen the stripes without diminishing the clarity of the image but nothing I tried work.  Finally I settle on an image that still had all the stripes it came out of the Fourier Transform Editor with, but the clearest image I had managed to achieve.  The tools I used to accomplish this after the Fourier Transform Editor were the Convolution tools Sharpen and Haze Reduction in ERDAS Imagine and in ArcMap I used the Spatial Analyst Focal Statistics tool with a width and height of 3 and a Statistics type of Range.

Intro to Electromagnetic Radiation (EMR)

Photo Interpretation & Remote Sensing - Mod 5a

This lesson had us work in ERDAS IMAGINE for the first time which was rather different.  We brought a raster image in and selected an area of Washington State to export to ArcMap using the Inquire Box and created a subset image.  After running the process we moved to ArcMap to manipulate the image properties and legend to display the information appropriately.  Below is the result.

Land Use\Land Cover Classification

Photo Interpretation and Remote Sensing - Mod 3

This week's lesson was to classify a TIFF image to Level II.  A polygon shapefile was created and two fields, Code and Code_Descr added to the attribute table.  Next an editing session was started and the boundaries for the different classifications needed to be digitized.  Digitizing each of those islands was a very long, drawn out process, but the most difficult part was determining how to classify everything.  So many things could be either one thing or another.  Fortunately most of what I was uncertain about fit into the same category so I didn't have to make a final determination.  If I had had to drill down to Level III there were several things I wouldn't have known how to classify.  There were a couple buildings that could have been schools, hospitals or nursing homes.  Fortunately they all belonged to the Level II Commercial and Services category so I didn't have to decide. 

This lesson was pretty intense.  There was a lot of reading material to cover and so much information it was difficult to absorb it all.

Visual Interpretation of Aerial Photography

Photo Interpretation & Remote Sensing - Mod 2

This lab covered the methods and techniques used to visually interpret aerial photos through three exercises.  The first exercise focused on tone and texture.

  Five areas of the map were selected to represent different tones in a range from very light to very dark.  Polygons were drawn around these areas then converted to features so the attribute table could be edited to add the tone of each feature and its label added to the map.  Next, the same was done for texture, selecting five areas that ranged from very smooth to very course.

Exercise 2 covered shape and size, shadow, pattern and association.  Three objects were selected to represent each of these elements except association which only required two.  Instead of polygons markers were used to highlight features that were identified in each category.  The markers were also converted to features and each attribute table edited to include the name of the feature so it could be labeled.

The last exercise was a color comparison between the same image in True Color and in False Color IR.  Again markers of the selected features were converted to features to be edited and labeled.  This was done in a data frame with the True Color image, the colors of the selected features were noted then recorded in a table in the process summary.  Those same features were then examined in the False Color image and the difference in colors were also noted in the table.