Need help in multispectral image classification using SVM in MATLAB..

 

i am not able to understand how to separate my input vector(Set of features after feature

 

selection step) into training and testing data??

 

Can anyone help me or provide me code??

plz mail me code at : smriti1486@gmail.com

 

Thanks in advance

We won't do your homework, we won't do your work.

 

Post your problem\code and we will try to help you.

 

PS: before posting a new thread, read the forums rules

I extracted around 15 statistical features from Multispectral image. and reduced it to 3-4 features using k-nn clustering algo.

 

i need to perform classification using SVM.

 

How can i divide this into training n testing set?

 

i am not able to understand how this statistical features alone can classify my MS image.

 

Refer me some tutorial if any..

Generally, a classification is done in two general ways, either by rules .. or by statistical learning\approximation ...

 

here is a link to some references on Mulispectral Images Classification using SVM:

 

- www.intechopen.com/download/pdf/pdfs_id/9501

 

- www.cic.unb.br/~mylene/PI_2010_2/ICIP10/pdfs/0003641.pdf

 

- http://ieeexplore.ieee.org/xpl/freeabs_all.jsp?arnumber=1529435

 

- http://spie.org/x648.html?product_id=815783

 

- www.gisdevelopment.net/technology/ip/ma0561.htm

 

- paper.ijcsns.org/07_book/200809/20080933.pdf

 

.. good luck

Thnx.. it helped..

Can anyone provide me multispectral image along with its training image(reference image)??

 

or links to download..

I don't have any .. but

 

Statistical Learning needs hundreds of images to train your model !