Lab exercise 3, due midnight September 14, 2004
ES5053: Remote Sensing, UTSA

Student name: ______________

Get starting with ENVI

Objective: get starting some basic operations in ENVI such as displaying an image in different ways (gray and color), getting a pixel position and value, linking multiple images, calculating NDVI, band math, statistics, and input and output image.

(1)    Now you need to creat 2 new directories in your computer called c:\RS5053\lab3\Data where you need to copy all provided lab data to and c:\RS5053\lab3\MyWork where your works and results will be located. All lab data and your results can stay there for one week. So it is your responsibility to backup all your works and results after each lab. In the following Wendesday class (one week after), it is the responsibility of a student who is using the coputer (no matter whoever used last week) to delete all data from the previous lab. Then this student can create new lab directory such as Lab4, or just simple change Lab3 to Lab4. To actually copy data, you need to go to the class website, then go to Lab3 (for today's lab), you will see there are two files you need to copy: one is the image file ETMp27r40y01m7d21 that is actually store all digital numbers of an image (22M), the other one is the head file ETMp27r40y01m7d21.hdr that actually store the coordiante and projection information, it is a text file (828k). Now right click  ETMp27r40y01m7d21, click Save Target As, it will bring a new window to allow you to locate your c:\RS5053\lab3\Data directory in your local computer. Please be sure to put a "  " to the file (it looks like "ETMp27r40y01m7d21") for the File name, then click save. Do the same thing to copy the head file to your local computer.

(2) Now you will use ENVI to do your job. ENVI is an image processing program commonly used for image processing and analysis by remote sensing scientists. Click  “START” à ”Programs” à  You will see there are a number of programs and documents. In this class, you will mostly use the "ENVI 4.0" and "ENVI online manuals and tutorials". The tutorials are great materials for you to learn how to use the ENVI package. In this class, some of data may come from the tutorial data. You are encouraged to explore the tutorials by yourself. If you finish all of the materials, I am sure you will be an expert in using ENVI for image processing and analysis. Click on the ENVI 4.0, the program will be running. Click “File” à ”Open Image File”. This will bring you to the default data directory (c:\RSI\IDL60\products\ENVI40\data). In this directory, there are some datasets for you to freely use. You can practise later on. Now you need to find your work directory where your data are actually locating at. There is another way to setup your work directory (you will learn it in next lab). Now go to the directory c:\RS5053\lab3\Data where your image data is, and choose the file ETMp27r40y01m7d21, which is a subset of ETM+ image with path27 row 40 acquired on 7/21/2001 prepared for this lab. This brings up a small window (called Available Bands List) that lets you pick different bands of Landsat TM data to view. You can view images in Gray Scale or RGB Color.

Question 1: Explain the difference of Gray image and RGB color image.

(3)    Click “RGB Color”, then sequentially click on “Band 4 (0.8300)”, then “Band 3 (0.6600)”, then “Band 2 (0.5600)”. Now click Load RGB, and now you should be looking at a Landsat ETM+ image of San Antonio, TX.  By the way, the (0.xx00) numbers that you sequentially clicked on correspond to the center of a wavelength band (in micrometers) of the data. For ETM+ image, Band 1 is blue, Band 2 is green, Band 3 is red, Bands 4, 5 and 7 are near infrared, Band 6 is a thermal band (10.4-12.5 micronmeters), and is not shown here. If you get stuck at any point while using ENVI, the The Start ENVI Online Help is highly recommended.

(4) Now you should see 3 display withows (Main, Scroll, and Zoom). Getting familar with these 3 windows

Question 1: what are the main differences of the 3 windows? (if you have difficulty, please refer to either the online help or tutorials

(5) Trying to see if you can find where Loops 410 and 1604, UTSA campus, and Downtown are located. Click on “Tools” à“Cursor Location / Value” of the main display window. This window tells you the location (x, y in pixels), screen values, projection, coordinates in meters and lagitude/longitude, and real data values of the pixel that the cursor is pointing to. 

Question 2: Explain why this image is called a “false” color image.
Question 3:  Note the R: G: and B: values that the “Cursor Location / Value” window gives you. When you put the cursor over red pixels, what wavelength are the data largest? What objects, which are present on the ground, are responsible for these red pixels? why? (hint, see slides about absorbtion of Lecture 2). Why is there a region of relatively dark in the mountains (the very northwest part) of the image, do you know what objects are present on the ground?

(6) In the Available Bands List, change the band composition to band 3 (R) + band 2 (G) + band 1 (B), click Display # 1 at the bottom of the window, Click "New Display", click "Load RGB". Click Tools à Link à Link Displays à OK, which links two images. Point your mouse to any red rectangle of the two image diplays, click and move the mouse, and then release your finger.

Question 4:  what you see during the above operations? what color you see in the #2 diplay window while it is red color in the #1 diplay window? what color you see in the #2 diplay window while it is dark color in the #1 diplay window (northwest part)? why they are so different in #1 while they are looks the same in #2? why we call the bands 3+2+1 composition of TM image is a "true" color image?

(7) Now we will calculate a Normalized Difference Vegetation Index (NDVI). We will do it two different ways, first manually and then by automated means. Click on “Basic Tools” à ”Band Math” from the main ENVI window. Enter the NDVI algorithm and then click “OK”. (HINT: you can find good help on how to enter an algorithm by clicking “Help” à ”Index”, then entering “Band Math, example function”). You will want to use the float( ) function to convert all parameters in your algorithm to floating point numbers. After getting past this hurdle, identify your data bands (e.g. “Band 4 (0.8300)” or “Band 3 (0.6600)” or….) with the variables you named in your NDVI algorithm, then click “Memory”, then “OK ”. This will result in ENVI doing the NDVI calculation that you just entered and spitting the result out to memory (this memory file will be automatically removed after you close the ENVI, to keep them permanant, you need to output it to a output file. It is better to output to memory if you just want to see it or if the reulting image is only a middle status image not a final one that you would like to keep). The same calculation is automatically performed by clicking “Transform” à NDVI (Vegetation Index), using default parameters, and output the result to Memory. Open these two new images in two new display windows.

Question 5: Use the Link Displays tool you just learned to see if these two new images have exactly the same pixle value? what it is the value range and the mean?  (HINT: use the Statistics tool. You can find good help on how to use the tool by clicking “Help” à ”Index”, then entering "Statistics")

(8)  When we use the “Transform” à NDVI (Vegetation Index) to calculate the NDVI, if you change the "Output Data Type" in the NDVI Calculation Parameters window from Floating Point to Byte, output the result to Memory. Open this image to a new display window.

Question 6: Compare this NDVI image with the previous two. Tell the difference.

Question 7: Explain the difference between Floating and Byte data types, list and explain other data types in ENVI. (HINT: you can get all these answers from the ENVI Online Help you just used previously.