Lab - Introduction to ERDAS IMAGINE （Part II）

Image Info • Enables you to view and edit many elements of a raster image file (.img), including statistics, map information, and projection information.

• to access, click on the icon or right click on image name in contents and select metadata

Inquire Cursor • gives you individual pixel information by using a cursor which displays as a crosshair in the Viewer window • lists all of the available layers and indicates which color represents each individual layer • For every layer, each pixel has a value representing reflectance • Higher values indicate brighter reflectance • to access, click on the icon

Digital Number • Number that represents reflectance value • Assigned to each pixel • Also referred to as bit value (BV) or pixel value (PV)

DN.png

Skip Factor • Tells the computer how many pixel values to look at when computing statistics • For example if skip factor = o 1 = look at every pixel o 2 = every other pixel is sampled & used to calculate statistics))

Histograms

DNCurvesNormal.PNG

• This histogram tells you that there are 110 pixels with a DN = 127 and 20 pixels with a DN = 70

Bimodal Histogram • Infrared bands (such as band 4) are easily distinguishable from other bands because they are bimodal due to the drastically different reflectance of water and vegetation in IR

DNCurves.PNG

Copy the Lab07 folder from G:Fall-2015\G526\ . Right click on the Lab07 folder and choose copy. Navigate to a local directory, right click and choose paste.

Start ERDAS IMAGINE and open the tmdata.img using the methods learned in Lab5. Begin by looking at each band of the tmdata.img individually (Gray-Tone), and a false color composite if you wish — however, click on the No Stretch box when displaying. Use Raster  Panchromatic (or Multispectral) tab to look at all bands. By using the No Stretch, it is almost as if no statistics have been created for the data. You will now go through the procedure to “build statistics” for tmdata.img.

From the viewer menu, under the Home tab, click Met adata Tab (under the Information tab) and click on View/Edit Image Metadata. In the Image Metadata window that comes up, under the General tab, look in the Statistics Info section. This is what IMAGINE looks at to determine a stretch. Let’s pretend that we have no statistical information and we have to “build” or create the statistics.

Press the “” icon. This button goes through the data and finds the pixel values and compiles statistics. Look at the options, but you need not change anything. Use the Help button to find out more about the options. Close Help when done. Press Ok to generate statistics.

1. What does the “skip factor” refer to? (1point)

After we have compiled statistics, we can look at histograms of the data. This is a common and useful exercise in remote sensing work, and it is used to get a “feel” for the data. In the Image Metadata window, press the icon that looks like a histogram. With the cursor, grab the corner of the histogram window and stretch it out so it’s nearly as wide as the screen. Look at the histograms for all bands. In the Image Metadata window, change the selected band and the histogram should automatically be updated.

2. Using the Statistics Info, fill in the following chart: (7 points)

Look at all the statistics information/histograms and answer the questions below:

3. Find the pixel value (digital number) with the most pixels (points) for each band (look at the mode value). List what the pixel value is and the number of pixels (see the histogram tab and by putting your cursor over any pixel in the histogram, you get the pixel value on the x-axis and the frequency of that value on the y-axis): (7 points)

TM Bands

Why are the pixel values with the largest number of pixels not the same as the mean values found in #2? (3 points)

5. Look at the histograms for bands 4, 5, and 7. Describe the major difference between these histograms and the ones for 1, 2, and 6 (use your lab guide to help you answer this question). (3 points)

6. What is causing this difference? (3 points)

7. Close the histogram window and File  Close from the Image Metadata window. Redisplay tmdata.img, looking at the same bands individually (in Gray Tone) as you viewed in the beginning of this lab, only this time do NOT click on the No stretch box. Does the image look the same as it did in the beginning? If not, how has the appearance changed? (3 points)

Display tmdata.img again, but this time select the True Color display option and create a true color composite using bands 3, 2, and 1 (band 3 to RED, band 2 to GREEN, band 1 to BLUE).

When the image is displayed, press the Inquire Icon “+” on the Viewer menu to bring up the Inquire Cursor. To zoom in press the magnifying glass with the “+” in it. To get back to the full image, make sure you have clicked on the arrow cursor in the viewer menu, and then hold the right mouse button and select “Fit to frame”, or simply click the “Fit to Frame” button under the home tab.

8. Fill in the chart below. Use values from the “File pixel” column. The values in the “LUT Value” are the computer brightness values assigned to the pixel after the raw DNs have been temporarily stretched for display/contrast enhancement purposes. In the Inquire Cursor window, in the upper left, change the coordinate type to File instead of Map. Note the “File X and Y” coordinates change when you move the cursor. (7 points)

TM Bands

9. Compare these values to the histograms and see where the pixel values you observe in the Inquire cursor window fall in the histogram (if you could not answer question 4 before you may be able to do so now). Comment on the brightness values you have observed (I expect an in-depth answer here). Focus on one of the areas you looked at in #8. (10 points)

Close all viewers when you are done and DO NOT save changes.

Exit ERDAS Imagine, you’re done!