Photo Interpretation and Remote Sensing: Lab 4

Feature #1: Large water bodies

    Fig.1- Feature 1: Layout- True Color
  •  Upon analysis in the ERDAS Imagine histogram, the most prominent features of the landscape in question are a series of large rivers and offset streams dividing the terrain. Band 4 is the ideal layer for identifying contrast between land and water, and in this particular image there is a spike between pixel values 12-18 over water that is not present on land. Below is attached a picture of the layout (Fig.1) in True Color which clearly shows the distinction between the "dark" water body compared to the surrounding land. 
  • Also shown is the Layer 4 histogram (Fig.2) for the water feature, with pixel values spiking at 12-18, and the complete spectrum of pixels throughout the layer profile. (Fig.3)

Fig.2- Feature 1: Layer 4 Histogram
Fig.3- Feature 1: Complete layer pixel profile

Fig.4- Feature 2: Layout- False Color


Feature #2: Clear-Cut Forest

  • The following feature displays a small spike in layers 1-4 around pixel value 200 and a large spike between pixel values 9 and 11 in Layer_5 and Layer_6.
  • The histograms below exemplify the aforementioned respectively in layer 3 and 5. Clear-cut forests uphold all the specified criteria throughout layers 1,2,4,and 6 as well.  (Figs. 5&6). Layers 1-4 look very similar, and so do 5 and 6. 
  • Figure 7 shows the comparative pixel values throughout layers for Forested vs. De-forested land. 
  • The band combination used to clearly isolate clear-cut patches of forest was Red 6, Blue 3, Green 4- commonly known as False Color/Vegetative analysis. The forested area appears as bright green and the clear cut regions of bare soil appear vivid pink or magenta. 
Fig.5- Feature 2: Layer 3 Histogram

Fig.6- Feature 2: Layer 6 Histogram

Fig.7- Feature 2: pixels in forested vs. clear cut land


Feature #3: Heavily Sedimented Fluvial Channels


  • The prompt for the final feature was to look for a water-based feature where in certain areas, layers 1-3 to become much brighter than normal, layer 4 becomes somewhat brighter, and layers 5-6 to remain unchanged. Heavily sedimented water with high turbidity, near urban developments demonstrated these characteristics. 
  • The band combination used was Red- 4, Blue 2, Green 3, also known as Color Infrared highlighted the turbid waters as bright turquoise against the otherwise dark blue/green water. 
  • Figure 9 shows a comparison of pixel values in High-turbidity vs low turbidity water. 


Fig.8- Layout: low turbidity vs. high turbidity water bodies

Fig.9- pixels values throughout layers in low turbidity vs. high turbidity water bodies






Comments

Post a Comment

Popular posts from this blog

Module 5: Exploring and Manipulating Data

Image
 Module 5: Exploring and Manipulating Data  This module focused on the implementation of cursors and dictionaries within newly created geodatabases. We learned a lot of different functions and techniques applying to geospatial data. The most notable problem I encountered this week was compiling the search cursor queries into one script. One of the steps in the project was to create a Search Cursor for the cities layer to find the Name, Population (POP_2000), and Feature of all County Seat Cities. Initially, I could only get the cursor to run by dividing each of these categories into a separate cursor file, generating a long list of either city names, populations, and features, respectively. This was for all intents and purposes functional, but was practically illegible and a definite eyesore. By trial and error I was able to create separate variables, and vitally separate print row.getValue statements for each of my fields to make the same script generate all inquiries simulta...
Read more

GIS internship: Post #4

 GIS Day   This week, our internship cohort was assigned to organize a GIS day celebration to bring people together through shared interest in remote sensing and geo-data. Luckily, I already had something up my sleeve that fit the bill, so I'm wholeheartedly embracing the serendipitous coincidence without any remorse. I joined forces with the GIS staff at my university to put together a drone course on campus, courtesy of the kind folk at Gulf Coast State College (GCSC), with whom we connected via a GIS user's group conference a few months back. This course will teach participants how to collect, manage, process, and interpret a suite of aerial data with the added plus of learning how to fly a drone... very cool stuff. The application of remote sensing capabilities in the GCSC lab is emergency management focused, so we will likely be interpreting data through a public safety lens, which is a very interesting perspective, likely novel to most of us.   The course start...
Read more