Towards an automated classification method for ureteroscopic in vivo kidney stone images using machine learning techniques