Geog 430 - Remote Sensing

Lab 2

Crunch through Chapter 7 of the erdas tour guide (copies handed out or on the shelves by the door). It will teach you how to do a rectification. Don't do the subpixel co-registration section(I didn't copy it). Ignore anything in this and any other lab that refers to Unix.

However, these days you usually get imagery that is already rectified.

BUT (there's alway a 'but,' isn't there?), you will probably have to work using old aerial photographs. Basically, you scan 'em and recitfy 'em. That's what you're going to do this week. Typically, but not always, you will do a image to image rectification. However, again, not always. So - in this lab we're going to do things the long, hard way (maybe later in the quarter we'll do it the quick easy way).

In the geog430/lab2/ subdirectory on the data drive, you will find two ellensburg jpeg images which you will need to rectify!

• get a partner (do this lab in pairs)
• print them out and ID a whole bunch of spots (ground control points) which you think you can find in the field. Make sure at least 4 are in the area in which the photos overlap. Get about 20 for each photo (including the 4+ that overlap). Make sure the points are scattered throughout the photos.
• check out a GPS receiver (see me)
• head into the field and get the gcp data - use nad27 as your datum, UTM as your coordinate system. Units/meters.
• come back in and perform a rectification using what you learned in Chapter 6. - basically, you will click on your control points on the image and type in the UTM coordinates. No image to image type stuff. Also, importing your whole gps control point file is a real pain. To be honest, it's faster to just type 'em in. As soon as you get things typed in, SAVE! And then do another save as, so that you have your initial stuff in and saved, as one of the processes involved is deleting stuff.
• MAKE 100% SURE the autocorrect button (top left of the GCP table window) is OFF! Otherwise, erdas will move your points around and generally make a mess of things.
• Draw both side-by-side - how well do they fit? (load both into the same frame)
• Also, load the Quad sheet (DRG, tif format) (also in the lab2 subdirectory). See how well things overlay.
• As you go through, remember, there are two saves you must do to save your GCPs.

Writeup:

• Initial printouts showing GCP's (basically, your field maps)
• Erdas printout showing the GCP's that you actually used.
• Notes - what worked, what didn't, steps, etc. So that I could exactly replicate your work if I had to.
• Discussion of your final results - RMS error? How big are the pixels (in meters)? How well did the two photos fit after the rectification? How well did they fit the DRG?

Worth 10 points. Due monday, 17 April, 2006.