wrapped StarFeatureDetector and GridAdaptedFeatureDetector

Since there is no a standard functions cv::min and cv::max which has cv::Scalar argument, removed the performance tests, 
calling the functions cv::min and cv::max with cv::Scalar argument.
Instead of them two performance tests were created which call the functions cv::min and cv::max with 'double' argument.

Perf tests: method ::perf::TestBase::getDataPath made public static; added base CUDA detection into run.py script