Object Based Classification Cont.

 This lab is a continuation of the last lab. We are working with classifying raster images and then performing simple analysis based of those classifications. The first step of this lab was to take raster of County Park and classify it. Then we had to use the value count provided by the classification and use the pixel size to generate area for each class.

Then we took that already classified raster and broke the classes into two parent classes of permeable and impermeable to find their areas.  

Then we were provided a road and asked to provide an analysis of its condition. To accomplish this we classified the road into pavement, cracks, and vegetation. Once this was done we again used the count value to find the area of each class. Then we generate a map showing classes and their areas to shows the amount of cracks compared to the total pavement. 

 

Object Based Classification

     This week labs deals with classify rasters by using object classification inside of ArcGIS Pro. The instructor provided a demonstration on operating classification wizard and then asked us to create steps to perform object classification in the future.  My generate steps are listed below:

1. Use the Extract Bands function to better identify your object by selecting which bands to extract
1. EX vegetation
2. Open the Classification Wizard
1. Confirm Object based classification is selected
2. Select Default classifciation schema
3. Segment the Images
4. Confirm the desired
1. Spectral Detail
2. Spatial Detail
5. Create the Preview Segment
3. Training Samples
1. Remove the default classes
2. Create new parent classes
1. set values
2. set colors
3. Create specific classes
1. Select the class and use the polygon tools to train the ai
2. Collapse the polygons to create a single class
4. Select the Classifier (Support Vector Machine)
5. Run the Preview Classification
6. Run the Classification and set the output
7. Reclassify any errors
8. Finish the Classification Wizard

After completing our instructions we were ordered to classify a neighborhood into permeable and impermeable surfaces. 

Volumetrics

 This weeks lab focused on using the Surface Volume Tool to generate volume outputs alongside with understanding resampling effects on data. So first we took the mission area and clipped away the area that does not belong to the stockpile. Then we complete a multitude of resampling to 10cm and 100cm. Then we use the surface volume tool to find the volume of the pile at the non resampled pile, 10cm resampled, and the 100cm resampled pile. This was completed for three separate days of data. The table below shows that volumes calculated at each sampling size and on each data. The pictures below show generated layouts of the dates and samples sizes.  The meta data on each map indicates the data and sampling size.

Feature	Resampled 10cm Volume	Resampled 100cm Volume
0722	42,269	42.054
0827	83,888	83,667
0930	54,815	54,564

Raster Calculations

     This weeks lab focused on map algebra and utilizing it provide specific outputs. Additionally, the lab introduced the topic of resampling. The idea of resampling to create less precise but more useful information for analysis. For example, without resampling you will have much more variation inf surface level and it will be more difficult to establish a surface level for analysis. The first requirement of this lab was to generate a map depicting all of the flat areas inside of the mission area. To accomplish this goal we used the aspect analysis tool and the map calculator to find the area without a slope. 

   Then we were instructed to find all of the areas between the elevations of 233 and 234 to have an idea of water collection. 

Then we were instructed to find all the areas with a slope exceeding 30 Degrees. To accomplish this we utilized the aspect tool.

  

Then we were instructed to again show areas that exceed 30 Degrees but only above an elevation of 245 meters. To accomplish this we kept the aspect results from the last map and used map algebra to combine the limitations. 

 

 

 

Pix4D Processing and GCP Correction

     This weeks lab we worked with the Pix4D program in order to complete some post processing and to implement the idea of ground control point corrections. The first portion of the lab dealt with understanding projection and correctly importing the images into the program. Then we had to import the ground control points into the program with emphasis on projection and coordinates. After both the images and the GCP points were imported we had to input corrections for the GCP points to be accurate. This process required selecting each ground control point and then locating it inside of multiple overlapping images. The images below show the corrected ground control points on the mini box on the right side. 

GCP 1

GCP 2

GCP 3

GCP 4

GCP 5

GCP 6

GCP 7

GCP 8

   After the post processing was complete and the data was corrected using the ground control points we were instructed to generate a layout of the ground control points. To accomplish this task we utilized ArcGIS Pro to show the overall data along with detailed insets that show the location of each ground control point. 

Spatial Joins and Analysis

     This weeks lab built upon the analysis of the FAA certification database of last week. The purpose of this lab was to introduce spatial joins, hotspot analysis, and kernel density. This first part of the lab consisted of using spatial joins and comparing their outputs. We created spatial joins using zip codes, fishnets and zip codes again but using a clip functions. 

Zip Codes Spatial Join With Indiana Boundary

                      

Fishnet of Indiana     

Zip Codes of Spatial Join Using Clip

    Then we used the different types of spatial joins along with the use of Kernel Density and Hotspot Analysis to generate a number of maps.  

    We complete the same kernel density and hotspot analysis for the entire United States.

Attribute Data Management

     This lab provided us with a database of all FAA certification holders the United States based on their listed address. The purpose of this lab was to give us a better understand of manipulating tables inside of ArcGIS Pro and utilizing the attribute query tools. First, we were required to complete a table join to combine the tables. Then we were asked to create a multitude of maps by utilizing the attribute query function. The first map we were required to generate was depiction of all of the certifications. 

   The next map generated was to depict solely the pilot certifications.

The next map was more specific and depicted only commercial pilot certifications

The next map depicted part 107 certification holders.

The next map worked to limit the field by focusing on FAA certifications inside the state of Indiana. 

The final map tool the limited field of Indiana and the limited the attributes back to remote pilot certifications. 

 

   

.