X
PrevPrev Go to previous topic
NextNext Go to next topic
Last Post 31 Dec 2013 12:24 PM by  anon
Feature Extraction and attribute scatter plot graph
 1 Replies
Sort:
You are not authorized to post a reply.
Author Messages

anon



New Member


Posts:
New Member


--
31 Dec 2013 12:24 PM
    Hello, I am performing Example-based Feature Extraction on a RapidEye image, using the following settings: Segment scale level - 50 Merge level - 80 3 classes - Forest (44 training regions); Non-forest (126 training regions); Water (25 training regions) kNN classification with one neighbor, threshold 5 In the process of setting these parameters, I would like to be able to visualize 2d scatter plots of the attribute values for the training regions, in order to better understand the relationship of the attribute values and the classes. There is an example of this in the ENVI help file for kNN, but I do not see any functionality within ENVI to perform this. I could create the scatter plots after the classification using the output attribute datasets, but this does not seem very efficient within the classification process. Does anyone know if it is possible to view these scatter plots within the Feature Extraction work flow? If it is possible, any advice on implementation would be greatly appreciated. Thank you. Jamie

    MariM



    Veteran Member


    Posts:2396
    Veteran Member


    --
    03 Jan 2014 08:00 AM
    I believe you would first need to generate the attribute images and then you could use them in the 2D scatterplot. In Feature Extraction, the output of the attribute images occurs after you select your training data unless you choose the Segment Image only option, which you can choose to output the attributes.
    You are not authorized to post a reply.