Main Index | Search | New user | Login | Who's Online | FAQ

deflector
(Stranger )
10/29/11 11:24 PM
 GIS Problem

Hi guys,

I am studying for a GIS exam and would like if you could help me out.

It's an old exam problem that I'm not entirely sure of.

The question reads: The diagram shows the capture of roads using on screen digitising. The digitising movements are numbered in sequence from 1-9.

The questions are:

1. how many separate spaghetti lines were created?

I counted 14. 1-2, 2-3, 3-5, 5-8, 8-4, 4-8, 8-5, 5-3, 5-6, 6-9, 6-7, 7-9, 9-8, 8-9.

2. identify where nodes would be placed to make the lines topological.

I assumed here the end points of the polygon. So, 1, 2, 7, 5, 6. Then if the rest of the image was contorted it would have the same topology I think?

3. Identify where all possible over/undershoot errors may have occured.

I think it is where the lines have been retraced back over each other. So in this case, 4, 5, 3, 9, 8.

4. Assuming topology was built for this figure, how many individual lines exist?

This should be 6. That is, 1-5; 1-2; 2-3; 8-9; 5-6; 6-7.

5. if you were to perform on screen digitising from a remotely sensed image, what would you need to do to the image in order to digitise it from within a gis?

I'm a bit lost on this one. I think you have to convert it from a raster model to a vector. This question was worth quite a few marks so it needs more elaboration than this if it is correct. The lecture notes kinda skimped on this stuff.

If anybody could help me out I would really appreciate it. Also if you could recommend any websites which are a good primer on this stuff I would be greatful.

Thanks.

scrabblehack
(Stranger )
11/01/11 09:44 AM
 Re: GIS Problem [re: deflector]

1.  I would say there were 3 strands of spaghetti here:  1-3, 4-7, and 8-9.

2.  I would say at 8, 3, 9, and o, where o is the center of the figure.

3.  Typically overshoot/undershoot errors occur where there is a hanging line segment not forming a polygon.  So I would say 7.

4.  Potentially 11; 1-2, 4-8, 8-6, 5-3, 3-6, 6-9, 9-7, 2-o, o-8, o-3, o-9.  If each little square represents a different-valued polygon I would say it would have to be this way.

5.  You're probably right on this one:  group pixels of similar value, then take the outside shell of the pixels.

 Jump to *GISCafe* -----   GIS Education   Business Geographics   GIS Tools   GIS - Remote Sensing; Imagery   Data and Metadata   Intergraph's GIS Products   Autodesk MapGuide Board   ESRI Product Board (Arcview, etc.)   GIS Job Postings   Classifieds   GIS User News and Reviews