Reference no: EM132731404
GEOG 404-504 Digital Techniques for Remote Sensing - Old Dominion University
Assignment: Interpretation and Analysis of Multispectral Images and Image Enhancement
The objectives of the assignment are to learn basic skills of interpreting multispectral image bands, to understand the characteristics and applications of multispectral imagery, to get acquaint with the procedures for producing color composite images, and to understand basic process of image enhancement.
→ If you are using a non-ODU computer with ERDAS Imagine installed, please download the zipfile Assignment6.zip under Blackboard\Submissions and unzip it to your class folder GEOG404 (or 504) on your computer (assuming you've created the class folder when working on Assignment 1).
→ If you are using an ODU virtual computer via MOVE, download the zipfile Assignment6.zip under Blackboard\Submissions and unzip it to your class folder GEOG404 (or 504) under the network drive W (assuming you've create the class folder under the W drive when working on Assignment 1 via MOVE). Each current student is given 500GB for remote sensing/GIS usage only with W drive. You should have access to the same W drive whenever you are using an ODU (virtual) computer.
Attention: When you use ODU MOVE, please DO NOT save any assignment folders in any hard drive(s), e.g., C drive of ODU virtual computers. You will NOT have access to any data that you save in the hard drive(s) of a virtual machine after the virtual connection is terminated.
Our remote sensing/GIS students shall be able to access the same W drive when using any ODU virtual computers. However, if you are not able to access W with ODU virtual computer, please report the issue to the instructor Dr. Hua Liu immediately.
Part I. Get to Know Norfolk, VA
→ Launch ERDAS Imagine.
→ Open ...\Assignment6\Assignment6_Data\assignment6a.img in the blank Viewer window. This image is a Landsat ETM+ scene acquired on March 26, 2003, 15:29 (GMT). Note: Before clicking OK to add the image to the viewer window, click the Raster Options tab > check the box besides Fit to Frame.
→ Examine assignment6a.img and answer
Question 1: What is the projection information for the image (projection & datum)? How many bands does this image have? (Hint: You may want to check the image metadata)
Next you will create a subset from the Landsat image to cover City of Norfolk.
→ Zoom into Norfolk area (as shown in the figure below). Right click on the image > New AOI Layer. A new AOI layer appears in Contents.
→ Click AOI > Drawing. Use the Rectangle tool to draw a rectangle to fully cover City of Norfolk and parts of other cities, e.g., Virginia Beach and Chesapeake (see the figure).
→ Keep the current Viewer window open. In the top menu of ERDAS Imagine, click Raster
> Subset & Chip > Create Subset Image. The Subset window appears.
→ In the Subset window, add assignment6a.img as Input File. Name the output file norfolk.img. Save the output file in the Outputs folder. Click the AOI button > check the box besides Viewer > click OKs to create a subset image from the original Landsat scene. Use the following figures as reference.
→ Save the AOI file into the Outputs folder (Right click the AOI file in Contents > Save Layer). Name it boundary.aoi.
→ Open norfolk.img in a new Viewer window. Before you press OK to open the image, make sure the box besides Fit to Frame is checked in the Raster Options tab. (Ask yourself why.) Leave the rest setting as it is.
→ Observe norfolk.img and answer Questions 2-5 based on the following figure. Use Google map as reference if needed.
Question 2: What area of Norfolk is identified in Location A?
Question 3: What is the name of the water body in Location B?
Question 4: Name the road in Location C.
Question 5: Name the geographic feature in Location D.
Next you will investigate band combinations for better view of the study area. You will examine each reflected spectral band of norfolk.img.
→ In the top menu, click Raster > Mutispectral > Bands. You can change the band combination with this tool. Use the following figure as reference.
→ Create different band combinations and answer:
Question 6: Which band combination(s) is the easiest in interpreting each of the following features: water bodies, vegetation, transportation networks, and urban areas? E.g., RGB: Bands 2, 4, & 3. Provide the answer as table. Attention: You must provide the answer based on your observation, understanding, and analysis of the image. No additional references should be used/cited in answering the question.
Part II. Perform Image Enhancement
In this part of the assignment you will learn how to perform some typical spatial and radiometric enhancements in ERDAS Imagine. Some descriptions were cited from ERDAS IMAGINE Tour Guide.
→ In the top menu of ERDAS IMAGINE, click Raster > Spatial > Convolution. The Convolution window appears.
→ In the Convolution window, use lanier.img in Assignment6_Data as input file. Save the output file in your Outputs folder with name convolve.img. Note: It is not necessary to add the .img extension when typing the file name because ERDAS IMAGINE automatically appends the correct extension. Use the following figure as reference.
Next, you must select the kernel to use for the convolution. A default kernel library containing some of the most common convolution filters is supplied with ERDAS IMAGINE. This library is opened in the Kernel Selection part of this dialog.
→ From the scrolling list under Kernel, click on 3×3 Edge Detect.
→ Click on the Edit button in the Kernel Selection box. The 3×3 Edge Detect dialog opens.
For this assignment, you use the Kernel Editor to simply view the kernel used for the 3×3 Edge Detect filter. However, if desired, you could make changes to the kernel at this time by editing the CellArray.
→ Close the 3×3 Edge Detect window. Click OK in the Convolution window to run.
→ Create a color figure ONLY containing two viewer windows (lanier.img and convolve.img) side by side (make sure to use Fit to Frame to maximize the image display in each window). Name the figure Convolution saved in JPG format. Hint: You may use Windows Snipping Tool or Windows Paint software to create the figure. No noise is expected in the figure, e.g., software interface. Also answer:
Question 7: Discuss the differences between two images, lanier.img and convolve.img. Discuss the possible application of convolution process in a remote sensing project.
Attention: You must provide the answer based on your understanding of image enhancement. No additional references should be used/cited in answering the question.
→ In the top menu of ERDAS IMAGINE, click Raster > Radiometric > Brightness Inversion. The Brightness Inversion window appears as shown in the following figures.
→ In the Brightness Inversion window, use loplakebedsig357.img in Assignment6_Data as input file. Name the output file inverse.img in your Outputs folder. Under Output Options choose Stretch to Unsigned 8 bit and Inverse. Use the figure above on the left as reference. Click OK to start the process.
→ In the Brightness Inversion window, use loplakebedsig357.img in Assignment6_Data as input file. Name the output file reverse.img in your Outputs folder. Under Output Options choose Stretch to Unsigned 8 bit and Reverse. Use the figure above on the right as reference. Click OK to start the process.
→ Create a figure ONLY containing two viewer windows (inverse.img and reverse.img) side by side (make sure to use Fit to Frame to maximize the image display in each window). Name the figure Radiometric saved in JPG format. Also answer:
Question 8: Discuss the possible application of brightness inversion and reversion process in a remote sensing project. Attention: You must provide the answer based on your understanding of image enhancement. No additional references should be used/cited in answering the question.
Finally, at the end of the assignment, submit the following to Blackboard under Submissions:
- a single PDF file (≥ 1 pages) including typed answers to all the questions and two maps Convolution (in color) & Radiometric (no other file formats will be accepted. no WORD document).
Attachment:- Digital Techniques for Remote Sensing.rar