GEOG
4/582 Lab 2: Knowledge-Based Classification
Introduction
In this exercise you will use the ERDAS Imagine Expert
Classifier to combine GIS and remotely sensed data to extract thematic classes
and conduct spatial analysis.
- Open
the ERDAS Imagine Online Field Guide Volume 2 and read the section on
Expert Classification (pages II-135 – 138). The ERDAS Online
Document can be found on the Windows Start menu. Ask the instructor if you
cannot find the documents.
- We
will do two Imagine Expert Classifier exercises on the ERDAS Imagine
Professional Tour Guides: I. Create a Knowledge Base (pages 215 –
230) and II. Create a Portable Knowledge Base (pages 230 - 251). You can download the tutorial file (lab02.zip) here. You will need to unzip the files to your
workspace in the C:\Users folder before you start the exercise.
- Use
the viewer to view the image files that will be used in this lab. This
will give you a better understanding of the area being studied. Complete
the tables in the deliverable item #1 below.
- Complete
the tutorial exercise.
- Modify
the rules in Exercise I (Create a Knowledge Base) to add a new hypothesis that
uses the TM Band 4 variable to identify water. You will need to find the
threshold value on the lanier
image. Save a screen-shot of the rule props dialog window of this rule and
produce a map of the classification. Insert both images into your report.
Deliverables
- Record
what the numbers represent in the input.img file
in the table below. You can find the information from the raster attribute
dialog window.
|
Class Number
|
Class Name
|
|
1
|
|
|
2
|
|
|
3
|
|
|
4
|
|
|
5
|
|
|
6
|
|
|
7
|
|
|
8
|
|
|
9
|
|
Lanier.img is a LandSat TM image
with 7 bands (layers). Use any reference you have to complete the table below.
|
Band (Layer)
|
Spectral Resolution (µm)
|
Band Description
|
|
1
|
0.45 ~ 0.52
|
Blue
|
|
2
|
|
|
|
3
|
|
|
|
4
|
|
|
|
5
|
|
|
|
6
|
|
|
|
7
|
|
|
Record what the numbers represent in
the supervised.img file in the table below (Note: you
can provide a screen shot of the attribute dialog window showing those class
names).
|
Class Number
|
Class Name
|
|
1
|
|
|
2
|
|
|
3
|
|
|
4
|
|
|
5
|
|
|
6
|
|
|
7
|
|
|
8
|
|
|
9
|
|
|
10
|
|
|
11
|
|
|
12
|
|
|
13
|
|
|
14
|
|
|
15
|
|
|
16
|
|
|
17
|
|
|
18
|
|
|
19
|
|
|
20
|
|
- A
screen shot of the rule props dialog window of the rule you created to
classify water and a map showing the classification result using the
modified knowledge base.
- Write
a short report describing the knowledge you used to classify images in the
first exercise (i.e., create a knowledge base),
including the modification you did to classify water.
- Use
your own words to describe how the “Flat Solid Open Ground
(go)” class was determined in the cross-country travel analysis.
- Use
your own words to describe how the “Closed Forest
(no go)” class was determined in the cross-country travel analysis.
- Produce
a map of the cross-country travel analysis. Describe what information is
on the map and how that map could be used in travel planning.