You are on page 1of 19

140920 FM-BINUS-AA-FPT-66/R7

Practicum Case
COMP6064 | COMP6590 | T0243
Geographical Information System
Computer Science O211-COMP6590-SL01-09

Valid on Odd Semester Year 2020/2021 Revision 00

Learning Outcomes
 Explain concept, component and the advantages of geographical information system (GIS)
 GIS concept and data model
 GIS data
 Geodatabase
 Spatial analysis using GIS tools
 GIS data standards and future trends

Topic
 Session 09 – Raster Analysis

Sub Topics
 Image Clustering
 Image Classification

Page 1 of 19
140920 FM-BINUS-AA-FPT-66/R7

Soal
Case

Raster Analysis

Raster Analysis is a tool to analyze part of a map and determine what the type of the place. In
ArcMap, raster analysis divided by 2 which are:
a. Supervised Image Classification:
Uses the spectral signatures obtained from training samples to classify an image. With the
assistance of the Image Classification toolbar, we can easily create training samples to
represent the classes we want to extract. We can also easily create a signature file from the
training samples, which is then used by the multivariate classification tools to classify the
image.
b. Unsupervised Image Classification:
Finds spectral classes (or clusters) in a multiband image without the analyst’s intervention. The
Image Classification toolbar aids in unsupervised classification by providing access to the tools
to create the clusters, capability to analyze the quality of the clusters, and access to
classification tools.

To do supervised image classification, follow the steps below:


1. Prepare a satellite map image and import it into ArcMap

Figure 1. Satellite Map Image

Page 2 of 19
140920 FM-BINUS-AA-FPT-66/R7

2. Enable ‘Spatial Analyst’ extension in extensions menu and ‘Image Classification’ toolbar

Figure 2. Extensions Menu

Figure 3. Extensions Panel

Figure 4. Image Classification Menu

Figure 5. Image Classification Toolbar

Page 3 of 19
140920 FM-BINUS-AA-FPT-66/R7

3. Use ‘Draw Polygon’ tool to take training sample. For this case, we divide the sample to 4
type: agricultural, uncultivated, water body and neighborhood. So first we take the
agricultural sample. The more training sample, the more accurate the data.

Figure 6. Agricultural Training Samples

4. Open ‘Training Sample Manager’ and merge the same type sample.

Figure 7. Training Sample Manager

Figure 8. Training Sample Manager Panel

Page 4 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 9. Merge Training Samples

5. Repeat step 3 and 4 until all type covered.

Figure 10. Uncultivated Sample

Figure 11. Merge Uncultivated Training Sample

Page 5 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 12. Water Body Sample

Figure 13. Merge Water Body Training Sample

Figure 14. Neighborhood Sample

Page 6 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 15. Merge Neighborhood Training Sample

Figure 16. Final Training Sample

Figure 17. Final Training Sample

Page 7 of 19
140920 FM-BINUS-AA-FPT-66/R7

6. Use ‘Interactive Supervised Classification’ tool and we will get the result.

Figure 18. Interactive Supervised Classification Tool

Figure 19. Classified image

Page 8 of 19
140920 FM-BINUS-AA-FPT-66/R7

To do unsupervised image classification, follow the steps below:


1. Use ‘Iso Cluster Unsupervised Classification’ tool

Figure 20. Iso Cluster Unsupervised Classification Tool

2. Input number of classes, in this case we try to input 4 first.

Figure 21. Iso Cluster Unsupervised Classification Configuration

Page 9 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 22. Classification Result

3. The result is not accurate because sometimes few different color is detected as the same
if the number of classes is too small. So, this time we try to use 20 number of classes.

Figure 23. Classification Result

4. Change the symbology variation to the same variation, for this example, we use green
for agricultural, blue for water body, sun for uncultivated, and rose for neighborhood.

Page 10 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 24. Changed Symbology

Figure 25. Map After Symbology Changed

5. Open ‘Search’ tool and find ‘Reclassify (Spatial Analyst)’ to merge the same symbology.

Page 11 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 26. Reclassify Search

6. Change the value to 1 for water body, 2 for agricultural, 3 for uncultivated and 4 for
neighborhood.

Figure 27. Reclassify Data

Page 12 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 28. Map After Reclassify

7. We also could convert from raster to polygon data by using ‘Raster to Polygon’. After
that, we could also set the symbology based on the grid code.

Figure 29. Raster to Polygon Tool

Page 13 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 30. Raster to Polygon Tool Configuration

Figure 31. Raster to Polygon Result

Page 14 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 32. Symbology Setting

Figure 33. Polygon with Symbology

8. We also could classify the grid code and store it into a new field, so we could symbolize by
the class name

Page 15 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 34. Add New Field

Figure 35. New Field Configuration

Page 16 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 37. Field Calculator

dim defClass
if [gridcode] = 1 Then
defClass = "Water Body"
elseif [gridcode] = 2 Then
defClass = "Agricultural"
elseif [gridcode] = 3 Then
defClass = "Uncultivated"
elseif [gridcode] = 4 Then
defClass = "Neighbourhood"
end if
Figure 36. Field Calculator Formula

Page 17 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 38. New Table

Figure 39. New Layer Symbology

Page 18 of 19
140920 FM-BINUS-AA-FPT-66/R7

Figure 40. Layer After New Symbology

Page 19 of 19

You might also like