ERDAS IMAGINE 2011 USER MANUAL

Prepared By:
Er. Sanjeevan Shrestha & Er. Shrijana Sainju
Instructor
Land Management Training Centre

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 1
 Exercise I
Introduction
 Start ERDAS Imagine by click Start/Programs/ERDAS 11/ERDAS IMAGINE 11.
You will see the main ICON PANEL.

 Click File > open > Raster Layer from main menu to display the image file to
be viewed. The Select layer To Add dialog box opens:

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 2
  In the select Layer To Add dialog, browse to your personal directory and select
sp28jul02rdens.img.
 Click the Raster Options tab at the top of the dialog, and then set the Layers to
Colours to 3,4 and 1 (red, green, and blue, respectively). Here, layer 3 is the
NIR band, layer 4 is MIR band 4 and band 1 is the Visual Green band 1.
 Click the Fit to Frame option to enable it

 After selecting proper display parameters, click ok, then image is displayed in the
viewer.
 Play with different tools such as zoom, pan, zoom to extent, Fit frame to
windows etc.

Inquire:
 To inspect the pixels, select Inquire cursor

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 3
  Drag the crosshair over the image and read out the values in the column “FILE
PIXEL” (the actual value in the image data file) and “LUT VALUE” representing
the value on screen for the bands that were selected.
 Try inquire box, inquiry colour, inquiry box etc

 Click on the Metadata icon on toolbar to show the information on the image

Task I: Investigate all information of the image

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 4
 Measurement Tool:

 Click on measure tool on toolbar, Measurement Tool dialog box

appears.

You can measure distance between two points, area under rectangular box, ellipse,
circles etc.

Link Viewer

 Click on Add Views > Create New 2D Views to open second 2D Views.
 Select File > Open > Raster Layer from the main menu. In the Select Layer To
Add dialog under Filename, select topo34f.img.
 In order to investigate how the image corresponds with map, you can link the

two viewers with Link View Toolbar. In the viewer showing the topographic
map the corresponding area of the image window is identified.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 5
  To inspect the pixels , you can select Inquire toolbar and you will see the
crosses in both images.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 6
 Swipe:
 For this purpose, close 2D Views 2#.
 Select File > Open > Raster Layer from the main menu. In the Select Layer To
Add dialog under Filename, select topo34f.img.
 Now you can see, both raster image; sp28jul02rdens.img and topo34f.img at
same place overlapping each other.

 Click Swipe tool, viewer swipe dialog box appears.

 Now slide swipe position arrow bar from 0 to 100 and analyse the result.
 Try to use blend and Flicker options and see what is the difference by analyzing
the results.

Band Selection:
 You can change the selection to bands by selecting Multispectral menu, you
can see

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 7
  You can choose you own band combination or there is a provision of various type
of band combination naming True colour, False colour, Desktop RGB etc.

Profile:

 To start Profile tool, First click on Multispectral menu, and then choose Spatial

Profile
 Also choose Surface Profile and Spectral Profile and analyse the result.

Spatial Profile: The spatial profile Viewer allows you to visualize the reflectance
spectrum of a polyline of data file values.

Spectral Profile: The Spectral Profile Viewer allows you to visualize the reflectance
spectrum of aewer single pixel through many bands

Surface Profile: The surface Profile Viewer allows you to visualize the reflectance
spectrum of rectangular area of the data file values in single band of data. You can
overlay the wire frame surface with image.

You can use help menu for any queries.

Try to answer these questions?

1. What is the Spatial resolution of image?

2. what is the approx. area and perimeter of single pixel?

3. How many bands are available in the image?

4. What are the corner values of the image?

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 8
 5. What is this in the image?

6. Are DN values of any pixel same in all bands? If no Why?

7. What is the projection system, Spheroid and Datum of the image?

8. What is the range of DN value? Why?

9. What does histogram suggest?

10. Is there any relationship between earth surface and DN value?

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 9
 Exercise II

Georeferencing
1. Georeferencing Topo-sheet
 To Georeference scanned raster surveyed toposheet data. Open Raster
toposheet data into viewer , Right click on the screen.

 A New dialog box appears to Select layer to Add, then select 7f2.tiff file,
click ok to open.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 10
  On erdas imagine software ribbon, click on Multispectral menu, click on
Transform & Orthocorrect > Control Points to provide Geometric model

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 11
  Select Polynomial as on set Geometric Model on Model List and Click
ok.

 GCP Tool Reference Setup dialog box appears to collect Geographical

reference point source, select the option Keyboard Only for the toposheet
map, then Click Ok.

 Click on Reference Map Information, click on set to get the Projection

Chooser dialog box
 On Projection Chooser window, Click on the Custom Tab and set
projection parameter as

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 12
  And then ok.

 Polynomial Model Properties dialog box appears, click Close to

proceed.
 Verify that the Set Automatic Transformation Calculation icon is off.

 Click Create GCP icon on Multipoint Geometric Correction toolbar .

Move the mouse cursor on the viewer, cursor will turn as „+‟ like cross sign
on the viewer, then take cursor to the top left corner of the grid on the
toposheet and click left button. GCP point gets placed at corner of
toposheet map grid then GCP point input values comes as in GCP Tool (X
input and Y input)
 For same GCP point input value for X Ref 77 and Y Ref 14 45 (to input
values use space bar eg. to enter 14 45‟, type 14 press space bar then
type 45).

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 13
  Follow the same above step to create atleast of second, third and Fourth corner
point on toposheet on other grid corner.
 Follow the same step for remaining another 12 GCP points on the toposheet grid.
 Click on Multipoint Geometric correction tool display resampling image
dialog, Resample dialog box. Provide TOPO-GEO as output name on your local
drive to take resampled output and make sure to ignore zero stats

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 14
  Remove previously opened toposheet non-georeferenced image from contents
by clicking right button to remove layer and open the resampled output topo-geo
image in same viewer.
 Use inquiry cursor to know exact location (where longitude and latitude values
are printed on Toposheet).

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 15
 Geo-referencing of Panchromatic Data

Objective: To Geo-reference imported pan Image with the reference of topo-poly geo-
referenced data.

 On erdas imagine software ribbon click on Panchromatic menu to get

respective toolbar.
 click on Transform & Ortho-correct toolbar control points to provide
Geometric model.

 Select Polynomial as on set Geometric Model or Model list and click Ok.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 16
  On GCP Tool Reference Setup, select Image layer (new viewer) and click ok.

 Dialog box appears to choose the reference image layer. Select topo-poly.img
and click OK. (This will be acting as reference image). Reference map
information dialog box appears and click OK.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 17
  As after providing three GCPs point, automatically prediction of point gets starts.
So place a point in one image, same GCP point will get placed in another image
automatically. If it is required adjust the same point to identified correct location.
 Collect sufficient number of GCP points and save it. Click GCP Tool menu file >
save input and file > save reference for the GCP points.
 Total RMS error can be checked by point respective GCP Tool Column.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 18
  Click on tool menu display resampling image dialog, on Resample dialog box,
write pan_poly as output name to get resample output and make sure to ignore
zero stats.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 19
 Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 20
 Exercise III

Image Subsetting
Objective: To Clip image area by using Area of Interest (AOI)

 Open Pan_rectified.img in a viewer

 click on Drawing menu and select Insert Geometry Rectangle button from
displayed toolbar to create rectangle AOI. So able to create Area of Interest by
digitizing the rectangle on the image.

 To subset image click on ribbon menu RASTER toolbar subset & chip sub
menu, Create subset Image to open subset dialog box appears.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 21
  In Subset dialog box make sure input as pan_rectified.img and output box
Type pan_sub.img. Check to Ignore Zero in Output Stats and Select AOI
Button in the subset dialog box to provide subset coordinate information.
 In choose AOI dialog box, select viewer radio button. Click OK in choose AOI
dialog box then OK in Subset dialog box.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 22
  When process get over, click on close button on process list and check the
subset layer by opening image in viewer by using quick launch open.
 Save the AOI layer by right click on content aoi layer Save layer as, specify the
layer name as study.aoi and click OK for future use.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 23
 Stitching of Images
Objective: Seamless Mosaicking of three Different images

In this exercise, different time period satellite LANDSAT MSS and TM overlapping images get
seamless Mosaic with each others.

The three input files to be mosaicked as: wasia1_mss.img, wasia2_mss.img, and

wasia3_tm.img.

 Click on Raster > Mosaic > Mosaic pro then new window comes up as Mosaic pro.

 In the MosaicPro Toolbar, Click on the Add Images icon, then Add Images for
Mosaic dialog gets opens.

 In the Add Images for Mosaicking under File tab, select Wasia1_mss.img, then select
“Image Area Options” tab and click Compute Active Area radio button to enable it
then click OK. The file wasia1_mss.img displays as image number 1 in the canvas of the
Mosaic Tool viewer.
 Repeat previous two steps adding wasia2_mss.img and wasia3_tm.img to the canvas
of the Mosaic Tool viewer.
 When the entire three images get added, the Mosaic Pro should look like the following:

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 24
  On image list, click on column visible to make images visible to visualize their
extent.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 25
  To Cutline Generation Options dialog box , select Geometry-based Cutline
Generation option then click OK to proceed. Images shown with new extent after
generation cutline.

 Click the „display color correction options dialog‟ icon in the Mosaic Tool viewer
. Color Corrections dialog box appears.

 Select „Use Histogram matching’ check box then click OK in the „color Correction‟ to
use function.

 Click on „Set Overlap Function’ icon

 The Set Overlap Function dialog opens to smoothing and Feathering function
distance 10 to apply function at this distance.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 26
  In the Mosaic Tool Viewer, Select Process/Run Mosaic, Output File Name dialog
opens.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 27
  In the Run Mosaic dialog under File tab, input name as wasia_mosaic in the local
directory then press Enter key button. Click the Output Options tab, enable the Stats
Ignore Value: 0 checkbox, then click OK.
 The job status dialog displays, showing the progress of the function, click close when
the job status dialog reads that the function is 100% complete.
 Open the stitched image in viewer to visualize it.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 28
 Resolution Merge
Objective: To generate Pan Sharp Image by using MSS and Pan Geo-corrected image.

 Firstly add multispectral.img and pan.img in the viewer.

 Firstly assess the metadata of both image especially pixel size and no of band.
Obviously, pan image has one band with lower pixel size (High resolution) while
multispectral image has several band with higher pixel size (Lower resolution)
 Click on ribbon menu Raster > Pan Sharpen > Resolution Merge to get resolution
merge dialog box
 Browse to provide input sharp image as pan.img and MSS image as
multispectral.img and browse to provide output name in local same directory resolution
merge.

 Select method as Principal Component radio button under Method, Cubic

Convolution radio button under Resampling Techniques, Ignore zero in Stats check
box under Output Options and type 3, 2, 1 in Select Layers the click OK to proceed.
 Progress Bar comes up, which shows the status of the job, click OK after completion.
 Open three 2d Views from Home menu window toolbar Add views subtool bar to
create new 2d views.
 Open Pansharp image in first 2d viewer, in second pan.img, and Multispectral image
in third individual viewer.
 Set the scale to 12500 by selecting Home menu Extent toolbar in each viewer
 To Geolink all three viewers, on Home menu window toolbar Link All Views select
Link All Views to link all viewers

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 29
  By visualizing all images, output Pansharp image have sharp color multispectral
resolution

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 30
 Exercise IV

Image enhancement
Objective: To enhance the image so that it would be easy to identify the features.

Enhance of image are divided into following categories:

1. General enhancement:
Here are going to see how resetting the value for the brightness/contrast help to
recognize the details that has been shaded before.
Add the panchromatic image to your viewer which is inside the folder resolution
merging.
2. Click on ribbon menu Panchromatic> General enhancement>
Brightness/contrast to get brightness/contrast dialog box.

3. Check the auto apply box and drag the cursor up and down to notice the
difference.
4. Drag the cursor manually to the extent that satisfy your requirement and apply it.
5. You can see that to some extent the shadowed area is brighten up and details
can be recognized.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 31
 6. Click on ribbon menu Panchromatic> General enhancement> histogram
equalize or Raster>radiometric>histogram equalization.
7. It will automatically stretch the radiometric resolution in a proportionate manner.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 32
 8. Similarly try the same with other option likewise and notice how the image is
enhanced.
Some of them automatically enhanced the image while other has to do
manually changing the value.

Spatial image enhancement

Filter operations are local transformation: a new image is calculated and the value of a
pixel depends on the values of its former neighbor. Filters are used for spatial image
enhancement , for example to reduce noise or to sharpen blurred images.
Convolution filtration: moving window method
1. Open multispectral img( inside folder resolution merging) in an emagine viewer.
2. Click on Multispectral menu> Filtering, the new convolve windows open.
3. On the left hand side of convolve window click on normalized kernel.
4. Then preview the enhanced image get visualized on right side, also click on the
label preview to get preview image label.

5. Now click on the Raster tab>spatial resolution>convolution

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 33
 6. From the above preview you can chose the required filter that serve best for your
post processing.
Note: high pass filters are used to sharpen the image while the low pass filters
are used for smoothing the image.
You can also explore with other option beside convolution like adaptive, non-
directional, crisp, texture etc and see how this work with the image.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 34
 Spectral Enhancement:
1. Normalized Differential Vegetation Index (NDVI)

 Open image 1999_1.img into viewer, then click on Multispectral Menu > Band
Chosser > False Color IR (FCC) i.e. Red color for NIR band, Green color for
Red band and Blue color for Green band.
 To calculate NDVI value of a multispectral image, click on Raster >
Unsupervised > NDVI, then new indices windows box appears.

 On new indices windows, provide the input file as 1999_1.img and output
name as NDVI_1999.img. Select sensor as Landsat TM and then select
function as NDVI, and then click OK.
 Open the NDVI_1999.img with raster options in pseudo colour image into the
viewer.
 Right click on contents > NDVI_1999.img > Display Attribute Table at the
bottom attribute table get displayed column as value, colour and other columns

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 35
 Note: NDVI values comes between -1 to 1 and positive value above 0.2 represents
vegetation on the image.

 Select the row which values comes over 0.2. It represents vegetation and click
colour column and provide a green colour.

Alternatively,

 Click on Toolbox > Model Maker > Model Maker , new model toolbox
appears.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 36
  Cilck on Place raster object and click it on model interface to locate it in
model.

 Again click on place a function and click it on model interface to locate it in

model
 Again repeat above first process.

 Now link raster object and function by using connect tool

The model should look like as below.

 Double click on first raster object, raster files opens.

 Place 1999_1.tiff image as input files and declare as integers.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 37
  Again double click on function on model, Function definition dialog appears
as:

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 38
  Change the function as other.

 In blank dialog, Write the function as:

(FLOAT ( $n1_1999_1(4) ) - FLOAT ( $n1_1999_1(3) ) ) /( (FLOAT (

$n1_1999_1(4) + FLOAT ( $n1_1999_1(3) ) )

Because NDVI = (NIR – Red) / ( NIR + Red)

 Click Ok.
 Again click on last raster object, navigate to your own folder and write name as
NDVI_1999.img. Select data type as Float Single. Click OK.
 Click Process > Run or click to run model
 Open the NDVI_1999.img with raster options in pseudo colour image into the
viewer.
 Right click on contents > NDVI_1999.img > Display Attribute Table at the
bottom attribute table get displayed column as value, colour and other columns

Note: NDVI values comes between -1 to 1 and positive value above 0.2 represents
vegetation on the image.

 Select the row which values comes over 0.2. It represents vegetation and click
colour column and provide a green colour.

Task: Prepare NDWI and NDSI image and analyse it.

NDSI = (Green – IR)/ (Green + IR)

NDWI = (NIR - IR) / (NIR + IR)

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 39
 Exercise V
Digital Image Classification and Accuracy Assessment

It is a process of assigning attribute value to the pixels. There exists a

relationship between land feature and the reflection values. In order to extract the
information from the image data, this relation should be found. The process of
finding relation is called classification. Or it is theme extraction process by sorting
out the pixels in different groups/classes.
While classifying image following steps should be followed:
Interpretation of satellite image – definition of classes
Field data collection
Create sample set
Classification
Accuracy assessment
1. The main purpose of a preliminary interpretation is to establish strata (classes
and
subclasses and their definition) for the selection of field data collection areas, as
described under fieldwork. In this exercise we will simulate field work by showing
terrestrial photographs of selected areas and describe the land cover and use.
2. We have to collect field data for image classification. Since we are not able to
collect data, we will just take some arbitrary data.
3. In this step, we have to create sample set for classification.
4. Classification

Unsupervised Classification

Objective: To classify image and determine land cover

 Display liss_sub.img image in a 2d view, click on Raster > classification >
unsupervised classification.
 Select Liss_sub.img under Input Raster File. Give output layer name
liss_class.img under Output File

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 40
  Type 8 in Number of Classes, Input 12 in Maximum Iterations and 0.950 in
Convergence Threshold, then click OK. Progress Bar comes up. Once it gets
completed then clicks close.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 41
  Display liss_class.img in a new View, by adding 2d view from home ribbon
menu window add views tool bar.
 Select Raster > Attributes in the viewer menu bar.
 Raster Attributes Editor which shows specified number of class and other
attribute.
Classes created as:
Unclassified
Class 1
Class 2
Class 3
Class 4
Class 5
Class 6
Class 7
Class 8
Colour as : Grey Levels
 Left click in box under color column and select the blue color
 Left click in box under Class Names and type Water
 Proceed for other classes to provide class name
 To add Area column, click on Table ribbon menu > Add Area and Add area
Column dialog box opens

 Select Hectares in Units and type Area in Name field then click ok. Area
Column gets added in Attribute Editor.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 42
 Supervised Classification
 Display image sp28jul02rdens.img in 2D view.
 Click on Raster > classification > Supervised > Signature editor. The
Signature editor dialog box displays.

 In signature editor, Select View > Columns. The View Signature Columns dialog
opens.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 43
  In the View Signature Columns dialog, right-click in the first column to access the
Row Selection menu. Click Select All.
 Shift-click Red, Green, and Blue in Column boxes 3, 4, and 5 to deselect these
rows.
 These are the Cell Array columns in the Signature Editor that you remove to
make it easier to
use. These columns can be returned at any time.
 In the View Signature Columns dialog, click Apply. The Red, Green, and Blue
columns are deleted from the Signature Editor.
 Click Close in the View Signature Columns dialog

Use AOI Tools to Collect Signatures

The AOI tools allow you to select the areas in an image to be used as a signature.
These signatures
are parametric because they have statistical information. For sample collection we
have to collect
ground samples. For that we can use GPS,Total Station or topomap etc.
 Use the Zoom in tool on the viewer toolbar to zoom in on one of the bright red
areas in the image file in the viewer.
 click on Drawing menu and select Insert Geometry Polygon button from
displayed toolbar to create polygon AOI. So able to create Area of Interest by
digitizing the polygon on the
image.
After the AOI is create, a bounding box surrounds the polygon, indicating that it is
currently selected.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 44
  In the Signature Editor, click the Create New Signature(s) from AOI icon or select
Edit > Add from the menu bar to add this AOI as a sig-nature.
 In the Signature Editor, click inside the Signature Name column for the signature
you just
added. Change the name to Maize, and then press Enter on the keyboard.

 In the Signature Editor, hold in the Color column next to Maize and select
Cyan. For evaluation purposes it is more suitable to select opposite colors from
the color of the cover on the image. In the end stage after classification the colors
can be changed for visualization purposes and map production. Under count you
see the number of pixels, which are sampled. In order to apply later on the
maximum likelihood classifier, theoretically you need n+1 pixels for each class (n
= no. of bands). It is recommended to have at least n x 10 pixels sampled, better
even 100 x n, by preference from several distinct areas.
 Zoom in on one of the black/dark blue areas in the sp28jul02rdens.img file in the
Viewer.
 Draw a polygon as you did in above 3 step. These areas are water.
 After the AOI is created, a bounding box surrounds the polygon, indicating that it
is currently selected. In the Signature Editor, click the Create New Signature(s)
from AOI icon or select Edit > Add to add this AOI as a signature.
 In the Signature Editor, click inside the Signature Name column for the
signature you just
 added. Change the name to Water, then press Enter on the keyboard.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 45
  In the Signature Editor, hold in the Colour column next to Water and select
White.
 Repeat this for all cover classes. At least add the light yellowish green areas in
the image. These are bare (agricultural) fields (e.g. photo 21 and 23 of the
Aerial photo show how such fields look in reality). Give it the name Bare and give
it the colour Red.
 Try other classes also and save that signature file.

Create Feature Space Image

 Select Feature > Create > Feature Space Layers from the signature editor
menu bar and feature space dialog box opens.

 Under Feature Space Layers, click the number 2 in the FS Image column in the
Cell Array to select the sp28jul02rdens_1_3.fsp.img row.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 46
  The output Feature Space image is based on layers one and three of the
sp28jul02rdens.img file. Layers one and three are selected since the selected
features are spectrally distinct in this band combination.
 Click OK in the Create Feature Space Images dialog to create the Feature Space
image for layers one and three of the sp28jul02rdens.img file.
 The Create Feature Space Images dialog closes, and then the Job Status dialog
box.

 After a process is completed, a viewer #2 opens, displaying the feature Space

image.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 47
 View Signature Object:
 In the Signature Editor menu bar, select Feature > Objects. The Signature
Objects dialog opens.
 In the Signature Editor, select the signatures for Maize, Water and Bare by
clicking in the Class row for Maize and Shift-clicking in the Class row for Water
and Bare.

 In the Signature Objects dialog, confirm that the Viewer number field is set for 2.
 Set the Std. Dev number field to 4.
 Enable the label checkbox by clicking on it.
 Click OK in the signature objects dialog. The ellipses for the maize, Water and
Bare signatures are displayed in the feature space viewer.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 48
 By comparing the ellipses for different signatures for a one band pair, you can
easily see if the
signatures represent similar groups of pixels by see-ing where the ellipses
overlap on the Feature Space image.
When ellipses do not overlap, the signatures represent a distinct set of pixels in
the two bands being plotted, which is desirable for classification. However, some
overlap is expected, because it is rare that all classes are totally distinct.
In this example the ellipses do not overlap, as only three and very distinct
features were chosen.

Task: View band separability for different objects in different band

combination before image classification.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 49
 Classify the image
 Firstly save the signature editor by clicking Ok. Give output name and location.
 Click Raster > Supervised > Supervised Classification to classify the
sp28jul02rdens.img. The Supervised Classification dialog box appears
 In the supervised Classification dialog, put input raster file as
sp28jul02rdens.img, browse signature file to locate previously saved signature
file and put classified image as supervised.img.


 Leave the Non-parametric rule popup list selected to None. And select Maximum
likelihood for parametric Rule.
 You do not need to use the Classify Zeros option here because there are no back-
ground zero data file values in the sp28jul02rdens.img file.
 A job status dialog displays, indicating the progress of the function. When the process is
100% complete, click OK in the Job Status dialog.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 50
 Accuracy Assessment
The Accuracy Assessment utility allows you to compare certain pixels in your thematic
raster layer to reference pixels, for which the class is known. This is an organized way
of comparing your classification with ground truth data, previously tested maps, aerial
photos, or other data.
In this example, you perform an accuracy assessment using the recoded output
thematic raster layer from your own supervised classification.
 Click Raster > Supervised > Accuracy Assessment to start Accuracy
Assessment utility. The Accuracy Assessment dialog opens.

In the Accuracy Assessment dialog, select File > Open or click the Open icon. The
Classified Image dialog opens.
 Select the file supervised.img from the list of files in the directory you just
opened.
 After that click on edit and create > add random points. Put number of points
10 and click ok.
 Click on reference column and give 1,2,3,4 …..as class values for each pixels
 In the Accuracy Assessment dialog, select Edit > Show Class Values.
 The class values for the reference points as collected from your classification
appear in the
Class column of the Cell Array.
 In the Accuracy Assessment dialog, select Report > Options. Check all options.
 In the Accuracy Assessment dialog, select Report > Accuracy Report.
 The accuracy assessment report displays in the IMAGINE Text Editor.

 If you like, you can save the accuracy assessment reports to text files.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 51
  Now your task is to analyze this report and get some ideas of image
classification.

Try to answer these questions?

1. Open your classified map and compare with topomap using swipe.

2. Use Arc GIS and prepare layout of this classified land use map and export as pdf.

3. Extract water bodies only from the image ?

4. Why do you think that areas near the river are also classified on buildup area?

5. Why there is small build up like area over water channel ?

6. what is draw back of pixel based image classification?

7. Try to classify different images.

8. What are the things to be considered to classify image accurately?

9. What is the problem of classification of image of higher resolution?

10. What does feature space suggests?

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 52
 Exercise VI

Vectorization
-47-ERDASine Customer Education
Objectives: Creating of GIS vector layers, editing and query making on it.

1. Open Erdas Imagine view and Open Raster Layer subset_quick

2. to create new vector layer click on application file> new> 2d view vector layer.

3. Create a new vector layer dialog box get appears Select Shape file(*.shp) in File
of type, Type Roads in File name
4. Click OK. New Shape file Type Option dialog box get opens.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 53
 4. Select Arc Shape under Select Shapefile Layer Option then click
OK.
5.Vector roads.shp layer is created on image and to use vector tool for Vectorization
click on ribbon drawing menu toolbar insert geometry polyline.

Adding Attributes to Existing Table

1. Click on ribbon Table menu toolbar query column properties. New column
attributes dialog box appears new button is used to create new column.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 54
 Editing the Existing Vector Layer
1. To make layer editable, click on drawing menu modify toolbar Enable Editing.
2. By using Drawing Menu insert geometry tool creation and editing of features can be
done.

Creating Attribute based Symbology

1. Click on format >vector symbology toolbar. Symbology dialog box get open

2. Select Automatic/Unique Value in the Symbology dialog box.

4. In the pop up list select unique value on the drop down list. (by using road attribute
column new styles are get generated).
5. Enable „Generate New Styles‟ check box and click OK.
6. In the Symbology dialog box, different color for different id put as row wise in shape
file.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 55
 Also able to change color, thickness and line style for each line.

20. We observe different line styles in Symbology dialog box.

21. Polyline in the shape file has been changed to different line styles and colors based
on attributes.
22. Click Close in the Symbology dialog box then attention messagebox appears click
Yes to save line style, color.
23. Appears Save Symbology As dialog box.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 56
 24. Go to the directory where shape file exists and type roads and press enter key
(*.evs files should have the same name as the coverage name, and it has to stored in
the same directory where coverage exists).

25. Click OK and close all windows.

26. Save symbology .evs file ensures to open shape file road with thesame style and
symbology

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 57
 Exercise VII

Map Composition

Objective: Creating map composition layout for printing

Before adding map composition view, open the set of files which is to be composed in
the view.

 On erdas imagine software ribbon, click on Home menu > Add View > Map
View.
 Open classified image in to 2D view.
 Click on layout menu to select A4 size template from toolbar.

 To Adding the image on Map view, click on 2D view and drag into map view
frame, so the map comes on to the frame.

Create Grid

 To create grid on to map frame, click on Layout > Insert Map Element Map
Grid toolbar.
 Click on the image inside the map frame on which coordinates grid need to
place. Set Grid/ Tick Info dialog box appears.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 58
  Click on Set destination Projection to open Projection chooser dialog opens.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 59
  Choose the following coordinate parameters.
Projection Type Transverse Mercator
Scale Factor 0.9999
Longitude of Central Meridian 84
Latitude of origin of Projection 0
False Easting 500000
False Northing 0
Spehroid Name Everest 1830
Datum Name Everest 1830

 Click OK.
 In the set Grid/ Tick Info dialog box, Select the “Meter” option under Map Units.
 Specify the 5000 m under spacing.
 Enable Use full Grid. Click on Copy to Vertical, this option will apply the same
setting to the vertical axis.
 Click apply and then click close.

Create Legend

 Click on Layout Menu > Insert Map Element Legend toolbar, then legend
instruction come up to select map frame by using left click.

 New dialog box for Legend Properties open up and provide name input then
click Ok.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 60
  In Map View, Legend Comes up with legend information.

Add a Map Title

 Click on Drawing menu > Insert Geometry > add text element
 Move the cursor to the top of map in the Map composer viewer, the cursor
becomes an I-beam.
 Use I-beam point to Enter Text String. Type Forest Type Map, and double click
to finish the text.
 The text is now displayed in the map composition.

Change Text Style

 Click on the text string in the map View to select it.

 From the Drawing menu styles toolbar, click on nudger to get Text style
Chooser, Select style, size and its unit. Then click ok on selected style.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 61
 Insert North Arrow

 Click on Layout Menu > North Arrow > North Arrow Properties toolbar, then
north arrow properties dialog box appears.

 Select style on north arrow button, symbol chooser dialog box appears.

 Click Standard menu and choose North arrows in the dropdown menu.
 Click apply and OK.
 Now, Click on Layout Menu > North Arrow > North Arrow and click on the
area where you want to place north arrow preferably at top right corner.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 62
 Scale Bar

 Click on Layout Menu > Insert Scale Bar toolbar, then legend instruction come
up to select map frame by using left click.

 New dialog box for Scale Bar Properties open up and provide name input then
click Ok.

 Tick on Meter-Meter and take 3.0 as maximum length. Click Apply and then
close.

Print the Map Composition

 Click on the File Menu > Print, then it opens save map view composition comes
up.
Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 63
  Save the map view composition to save layout.

 Click on the Print Destination popup list to select the Image File.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 64
  Browse on the plot file and give suitable name in your working folder.
 Now Select page setup tab, Select the number of panel as 1. The number of
panels set to 1 means that the total map composition will be coming under single
image.
 Click OK.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 65
 Exercise VIII
Forest Cover Classification
1. Open ERDAS IMAGINE and select the File Tab.
2. Click the Open tab and select Raster Layer.
3. Next, navigate to the folder on your hard drive that contains your imagery. In the
case below we are opening a 5 band multispectral Ortho Ready Tiff of Rapid
Eye. Select the image file and then click OK in the window.

The image will then open and automatically load the True Color bands into the Red,
Green and Blue guns. Under the Multispectral tab up in the menu under the Bands
section the user can load different band combinations. In the drop down select Rapid
Eye (See Fig.3). You can pick different band combinations, e.g. true color R,G,B, False
Color IR etc.. For the example below we selected the False Color IR to select our
Supervised Classification samples.

Fig.3. Multispectral Tab (True Color)

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 66
 Fig.4. Multispectral Tab (True Color)

4. Under the Home tab select Fit to Frame and the image will zoom out to its full
extent.
5. Similarly add the vector layer provided. In this case you are provided a point
layer (GPS points) containing the species of forest.
6. Open the attribute table of the vector layer and see the CLASS field.

AOI tool
Once the satellite image is loaded the next step is to start collecting samples for
classification. This is done using the AOI tool.
1. To create an AOI go to the File tab > New > AOI Layer.
2. Open the AOI tools.
3. Next, choose Signature Editor from the Classifier button menu in the main
Erdas toolbar.
4. Click on the polygon button to insert a geometry.
5. To Highlight the point of interest whose signature file is to be created, select the
row containing similar species class. (For e.g. Hill Sal).

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 67
 6. Click the Vector Tab > Viewing Properties, properties dialogue box appears.
Check the Display selected Subset Only box.

Fig. 5. Viewing Properties

7. Zoom into your first area to collect a sample using the zoom button and begin
collecting your first sample.

8. To collect a signature, choose the Create Polygon AOI button from the AOI
tools. Digitize a polygon around a sample forest species, and double-click when
finished.

9. Click the Create New Signature from AOI button in the Signature Editor to
add the sample. Repeat this for at least five forest species of same category.

Five or more signatures should be collected for every land cover type to be
classified. Also check the pixel count number in the Count column of the
Signature Editor. Signatures should contain at least 30 pixels but not be
excessively large. Once you have a minimum of five signature of similar forest
species you can merge these into one to create one class taking into account all
five aoi’s spectral properties you’ve collected. See fig below:

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 68
 Fig. 6. Merge selected signatures

10. In the example above (Fig. 6) select all four classes by holding the shift key and
left clicking in the Class # column.

11. Next, in the menu, select Merge Selected Signatures button. A new Class will
appear.

Fig. 7. Merge signatures to create new class

In this example our new merged class is listed on the bottom and its default is Class
1.
12. Double left click on this new Class1 and rename it to something representative of
the class. E.g. Hill Sal. You can then delete all the other classes that were input.
13. Repeat this process for the other classes from the aoi.

Once you have your signature file with all the classes you would like to identify. You
need to save this signature file.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 69
 14. In the Signature Editor Window click on file and save as. Navigate to the folder
you want to save the signature file to and give it a name. Then click OK

Supervised Classification
The next step is to open the Supervised Classification window.
15. Under the Classifier tab click on Supervised Classification. A Supervised
Classification window will open.
In the Supervised Classification window select the input raster file. This should
default automatically to the imager you have opened in the main viewer. If not click
on the open folder and navigate to the 5 band Multispectral image you’ll be
classifying.
16. Next select the signature file you just saved. Open the folder next to the input
signature File and find the .sig file you saved.
17. Select the folder you want to save the output classification file in, in the
Classified File drop down. Navigate to a folder on your drive and give it a name.
This will be an .img file.

For this example we will use the default parameters and run the classification using
the Parametric Rule of Maximum Likelihood Supervised Classification method.
18. When you are ready to run the classification Click OK.
19. You will see a process window running the classification. When done click
Dismiss

A Classification .img will be output to the folder you’ve selected

20. To view your classification file navigate to this folder through the File > Open >
Raster Layer. Select file type of .img and select your file. Click Ok. It will then
add to your menu in the Viewer.

21. And finally, to change the colors of the classes to a better representation right
click on the classification.img in the contents window and open the Display

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 70
 Attribute Table. This will bring up the attributes on the bottom of the viewer. Find
your classes and right click under the color column and select the colors you
want.

Accuracy Assessment

1. First open the Satellite_image.img in a Viewer.

2. Click Raster > Supervised > Accuracy Assessment to start Accuracy
Assessment utility. The Accuracy Assessment dialog opens.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 71
 3. In this window, go to File > Open and select the image that resulted from your
supervised classification. Navigate to the file and then click OK to load it. This
will be that image that will be used in you accuracy assessment.

4. In the Accuracy Assessment window go to View > Select Viewer and then
click on the Viewer that is displaying the 9159916.img image.
5. Now go to View > Change Color and be sure that the Change colors window
looks like this:

"Points with no reference" should be white and "Points with Reference" should be
yellow. Click OK when this is done.

6. From the Accuracy Assessment window go to Edit > Import User-defined

Points. Select the file with your GPS coordinates from the ASCII Point File
window that pops up and hit OK. You now see an Import Options window.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 72
 7. For Field Type, select Delimited by Separator and don't worry about the rest
right now. Click OK.
8. In this window, go to View > Show All and notice that the points now appear on
the satellite_image.img image in your Viewer. If you Zoom In on the image, you
can see that these points more clearly.

9. Now you can enter your class values for the reference numbers in the Accuracy
Assessment table.
10. Go to Edit > Show Class Values and the Class column will be filled in with the
values that are the actual values of the pixels located at the X-Y coordinates that
you indicated.
11. Finally, you can look at the statistics of your accuracy assessment by first going
to Report > Options. Check to make sure that the Error Matrix, Accuracy
Totals, and Kappa Statistics options are all turned on. If they are not, do so.
12. Then go to Report > Accuracy Report to generate Accuracy table. Analyse the
table.

Pre
Prepared By: Er. Sanjeevan Shrestha and Er. Shrijana Sainju Page 73