What type of images are you dealing with and what kind of classification you want? Please tell us something more about so that we can make meaningful suggestions.
You may find many classifier for classification purpose in image processing.
After feature extraction next task is to representation of those images i.e creating knowledge base of those images. Finally you divide your data set into testing data set and training data set .
For training data what u hv already done is enough( feature extraction) , then match the images from u r testing class images to training class.
I Have images called birds i get it from the following link
http://www-cvr.ai.uiuc.edu/ponce_grp/data/
i want classify them using SURF , using any classification method (in matlab) my problem is i cant figure how can i construct test and train vector from Extracted Surf features.
Create a matrix for training data with number of samples(birds) as rows and its SURF features as columns. then you use this training data matrix into the classifier function.