before begining

#sift_dector is suitable for parallel
#I sew the discusion on openMVG'pull requests "https://github.com/openMVG/openMVG/pull/837",And it was refused by the master.
#The main reasion is show below.
#1.The CUDA is not portable which only suported by nvidia device.
#2.The performance of the cuda_Sift is different with the performance of CPU code which test on the
#  dataset of https://github.com/jieeeeeeeeeee/Features_Repeatability
#3.There are some bug on people'test which I gurss is the reason for the different device.
#4.The license cause some problem.