*Note: IDRISI commands are specified as follows:
Commands:  COMMAND 1 – COMMAND 2 – ETC.
This format applies to all remaining laboratory assignments.
(If you would like more detail on any commands or functions, refer to the IDRISI ‘Help’ utility)

Digital images are often stored in one of three formats – .BIL, .BIP, or .BSQ. These file structures are described below (Campbell, pg.100-103; Jensen, pg. 60):

BIP – “Band Interleaved by Pixel” – Data are organized in sequence of values for Band 1, Pixel 1; Band 2, Pixel 1; Band 3, Pixel 1… Band 1, Pixel 2; Band 2, Pixel 2…  Each pixel is represented in all four bands before the next pixel is encountered. (Campbell, Figure 4.12)

BIL – “Band Interleaved by Line” – Data are organized in a sequence of values for Band 1, Line 1; Band 2, Line 1, Band 3, Line 1… Band 1, Line2; Band 2, Line 2… Each line is represented in all four bands before the next line is encountered. (Campbell, Figure 4.13)

BSQ – “Band Sequential” – Data are organized in a sequence of values for Band 1, Band 2, Band 3… Each band is treated as a separate unit. (Campbell, Figure 4.14) (Jensen, Figure 2-44

Note:  Several variations of the “sequential” format are used – most notably in the case of  LANDSAT data – including HDF (Hierarchical Data Format), GEOTiff and FAST-L7. Most software packages, including IDRISI, provide conversion tools for
these formats as well.

To demonstrate the process of converting image formats, and to increase your understanding of how to images are digitally structured, a 1993 SPOT BIL image of Portland will be imported into IDRISI.

Importing Images into IDRISI

Most multi-spectral satellite imagery, whether from LANDSAT, SPOT, or other platforms, consists of at least two digital files in its original form: 

·        The actual image file(s), organized by band, line and/or pixel.

·        A “header,” or metadata, file containing information “about” the image, including the image date, size and projection.

Though some image formats allow the software to identify, process and use the metadata as part of their importing routines, in many cases – such as with converting a BIL image in IDRISI – it is necessary to enter this information manually. This provides a good opportunity to develop a better understanding of file structure and the various image parameters.

For this exercise, a monochrome, single band SPOT image of Portland, in .BIL format, will be imported into IDRISI.

Before beginning the import process:

Using Windows Explorer, open the LAB4 directory. The folder should contain four files:

SPOTpdx.bil    –       the .BIL band image.

SPOTpdx.hdr  –       the header file, or metadata.

SPOTpdx.stx  –       a specialized file for use with some GIS software.

SPOTpdx        –       a specialized file for use with some GIS software.

Open the metadata file in any text editor (such as Notepad). Review the header information and familiarize yourself with its contents. You are now ready to convert the image file.

STEPS FOR IMPORTING IMAGES:

1.      Set the IDRISI project environment to your LAB#4 directory.

2.      Commands: FILE – IMPORT – GENERAL CONVERSION TOOLS – BILIDRISI

3.      Enter the image file name – SPOTpdx.bil (or browse to the file.)

4.    Specify the output prefix (pdximage). Do not use the same name as the input image.

5.      Enter the number of bands.

6.      Accept all other default settings.

7.      Select “Output Reference Information.”

8.      Refer to the metadata’s “Image Information Block.” Input the number of columns.

9.      Input the number of rows.

10.  Input the Minimum/Maximum X and Y coordinates. This requires some calculations. The metadata gives you the upper-left coordinates (ULXMAP, ULYMAP) in meters (MAPUNITS), and the resolution is 10 meters (XDIM = 10, YDIM =10). Use this information, along with the number of rows and columns, to calculate the lower-right corner X/Y coordinate. (Hint: Each row and column is 10 meters wide. It may be helpful to draw a diagram.)   

11.  Refer to the metadata’s “Map Information Block.” Select the “Reference System” browse option to locate the c:/IDRISI32/Georef directory. Scroll down the list of projection options until you find the appropriate reference system for this data. Refer to the metadata field UTM_ZONE for the image’s UTM zone. (NOTE: The ‘N’ option is for areas in the Northern Hemisphere, ‘S’ is for areas in the Southern Hemisphere.)

12.  Accept the default reference units (meters.)

13.  Accept the default unit distance.

14.  Select ‘OK’ to record projection information.

15.  Select ‘OK’ to convert image.

Display the result with the Grey Scale palette with the “Autoscale” option. Select ‘Layer Properties;’

QUESTIONS:

1)      What is the area of this image in square kilometers?

2)      Given the metadata states that this image represents a 7.5° USGS quadrangle (7.5° by 7.5° square), why is the image significantly taller than it is wide?

3)      Would this also occur if the image were of the Panamanian coast? Why or why not?

4)      What is the approximate distance across the Burnside Bridge (in meters)?

5)      If this were a LANDSAT 7 Enhanced Thematic Mapper (ETM) image, how many bands would need to be imported (including the panchromatic image)?

6)      All remotely sensed images are based on a raster model of geographic space. Briefly explain how raster models represent the landscape. What are the benefits/costs of increased resolution?

 

 

7)      Import the file Pdxspotm  You will find this on the Department of Geography computers.  What is the resolution (pixel dimensions) of the image?

 

8)      Display all bands in the grey scale, with autoscale and title options activated.  In the layer properties, you may want to with brightness range.  To keep everybody more or less consistent, keep the minimum set at 0 and adjust the maximum.  If you decide to adjust the brightnesses, make sure the values of the range (0 at minimum; ?? at maximum) make sure they are exactly the same for each band.

 

9)      How many bands are there?  What is the wavelength range of each band?  Do the titles for each image match the actual band designation?  For one clue, look at the general brightness for each image compared to one another.  You will need some sense of geography of the region.  If the titles are not correct, what is the correct designation and briefly discuss how you arrived at your answer.

 

10)   Play with the Composite module under the Display pull down menu (next to File) at the top of the IDRISI window.  What is the purpose of the Composite module?  What three general landscape categories are obvious in your composite image?