classification problem is static hand gesture recognition for android smart phones.
my problem is I am new with openCV so I want to know if pca is the best feature extraction technique for this classification problem because I was really confused for the many feature descriptors in opencv such as hog and surf but based on research I concluded that pca achieves good results with svm so if it is the best for my case ,can anyone give an example on how to use it to get feature vectors?so i can use these vectors to train my svm model?and if it is not can you still give me an example on the best feature extractor to use with svm ?
Thanks in advance^^