Features selection for image clustering using limited device

Image classification is becoming a prominent field with the passage of time. Image classification on limited devices such as cellular phone is also coming in demand as the technology is advancing. So for constructing classifier the use of clustering is coming into picture for its ability to use less space which is a prime factor for limited devices. Thus for constructing the classifiers several features have to be considered. This paper shows what features can be considered in order to create an efficient classifier on limited device such as cellular phone.