Please note that a newer version of MultiSpec32 (03/08/00) has been released by the Purdue group. MultiSpec16 files can now be opened with the 32-bit version of the software, and many other bugs have been fixed. You can download this version from the MultiSpec Homepage at:
http://dynamo.ecn.purdue.edu/~biehl/MultiSpec/
For those of you that will continue to use MultiSpec in the classroom, it is to your advantage to keep up with all of the new software releases. MultiSpec 32 is designed to take advantage of 32 bit operating systems and is a much more powerful and capable version than MultiSpec 16. If you need help upgrading to the 32 bit version, just ask.
Note: For those of you who do not wish to upgrade, you should be able to complete your project with the MultiSpec 16-bit software.
Project Overview:
Your final project for this course will allow you to apply what you have learned in class to a specific problem of your choosing. We will spend the last 2 or 3 (if needed) lab periods working on your final term project. This should allow you plenty of time to prepare the images that you will need to work with. You primary goals for this class are to be proficient in the use of MultiSpec software, be able to select, load and manipulate image data, and to solve specific problems using multispectral imagery, above and beyond the levels of the GLOBE tutorials. You should take full advantage of the capabilities of the MultiSpec software.
Project Requirements:
First, you should clearly identify and define the specific problem that you wish to solve using your analysis of remotely sensed imagery.
Some practical ideas could be: 1) change detection of specific elements in two or more successive images, 2) classification of scene elements, including percentages and types of land use for an urban area, percentages of agricultural land vs. urban/undeveloped land, forest cover vs. non-forest cover, etc., 3) comprehensive description of image, 4) geological description of image, or 5) multiple combinations of the above.
You should describe the methods that you used to determine your descriptive factors (i.e. band combinations and their particular uses, classification schemes and their results, etc.).
The format of the report is up to you. A standard paper, with referenced figures, containing an Abstract, Introduction, Methods, and Results should suffice. Since this class is primarily concerned with imagery, your report should include developmental as well as final prints of your images utilized as figures for your report. The color printer in Lab 409 (DECO2) may be used for printing your report images and illustrations.
Obtaining your image:
To see a list of our available image data, visit the PACES website at
http://paces.geo.utep.edu/index.html
and view the available imagery list or map provided in the "What we have" section. To obtain path and row location information for any particular region, go to the following website:
http://edcwww.cr.usgs.gov/Webglis/glisbin/advanced_search.pl?LANDSAT
This website will allow you to click on a region of the globe and will return path and row information for the particular region. From there you can figure out if PACES has the image data in its inventory.
Preparing your image(s):
Prepare a standard 7 band multispectral image subscene(s) (of your choice) from image data. Limit your subscene to a rectangular area of no more than 600 lines by 600 columns, preferably smaller.
Once you have determined the correct path and row information for your image(s), see me or Dr. Keller to sign out an image data CD. There is a one week time limit on the image data. You should fill out the sign out sheet in the image cabinet. If you aren't going to use it at home or elsewhere, please check out the disk and return it after the lab. You should have enough time to prepare your subscene within one lab period. You should create a subdirectory on the machine that you are using to save your image files to in the /images subdirectory. Also, you might want to bring a zip disk so that you can backup your image data before you leave.
At this point, you will need the modified handout from lab 2 "Cookbook Image Data Loading Using MultiSpec 32 (dated 03/08/00). Please disregard your lab 2 handout from Feb. 02, as this modified "Cookbook" handout supersedes your earlier one. You can find a copy at
http://www.geo.utep.edu/pub/seeley/cookbook.html
Please remember not to save the entire scene, only a subscene of 600X600 pixels.
Image Analysis and Report:
At this point, the direction you take to solve your problem and report your data is essentially up to you. If you need help, please feel free to call me at 747-5507 or e-mail me at seeley@geo.utep.edu.