Evaluating classification results (part 2)

In this video you will see how to use contingency to quantifiy the quality of your training sets. in the end the first usage of the accuracy assessment tool is shown.

Category:

Science & Technology

Tags:

• erdas imagine
• separability
• classification
• how-to

[bozcaada1735]

Hi, your suggestion resolved my problem. Thanks a lot.

[ricckli]

@bozcaada1735 sounds good...

[bozcaada1735]

Hi, Your tutorials are super. Thank you so much. I have a question for you? I am working on accuracy assesment. I click "add random point" for example 200 points. But most of these points are out of my image. I try several times. But a few points on my image and rest of out the image. Have you an explain this situation? thanks a lot for your help

[ricckli]

@bozcaada1735 there should be a solution: you can choose several classes for the random points. check all classes except the "unclassified"! see for reference: page 475 in the TourGuide (search on google). Good Luck and thanks for your comment!

[dennisalfiom]

como hago para obtener este tutorial y otros mas que me interezan bastante

[ricckli]

@dennisalfiom Sorry, I don't understand spanish! otherwise I would help!

[shaziaiesec]

Thanks mate, your tutorials are very helpful. Am working on accuracy assessment and it's a real headache >.<