--- Start of Idrisi List message
G'day everyone,
I have an interesting problem (at least to me). I wish to share it with you
all and request you for help/ any hint/ leads on the following issue. 

On the basis of unsupervised classifcation, I got a classification accuracy
of about 92-95%.  This classification also showed the change of about 8% in
tree cover on a 5 yr monitoring cycle. This is a situtation where detected
change and the misclassification errors are of comparable/ reltively
significant magnitude. Reducing error through ground truth is not possible
due to limited (already used) historical data available.

The issues before me are:

1. What should I do in such a case? 

2. Does it mean that in such cases, it is not possible to reliably monitor
tree cover changes over small time periods? Any suggestions to deal with
this situtation?

3. Is it possible to construct some examples of what period would need to
elapse before the effects of misclassification became negligible relative
to real change? 

I am at a loss to think of anything else than blaming the lack of
sufficient historical data for ground truthing.

Many many thanks in advance for your help.

Cheers.
Vivek
Vivek K.Varma
Department of Forestry
Institute of Land and Food Resources
University of Melbourne, Vic. 3010
Australia

Phone: +(61) 3 8344 5238 (Work); +(61) 3 9383 3417 (Home)
FAX: +(61) 3 9349 4172
Web site: http://www.GeoCities.com/RainForest/Vines/5953/index.htm
--- End of Idrisi List message

IDRISI-L is an unmoderated list. As a service to list users, please post a 
summary of useful responses to your questions.
To post a message to IDRISI-L, send it to [EMAIL PROTECTED] To 
subscribe to the list, send an email from the address to be subscribed to 
[EMAIL PROTECTED] with "subscribe" (without quotes) as the first line 
of the message body. To unsubscribe, do the same, only with "unsubscribe" 
as the first line of the message body.

Reply via email to