I guess without using the specific eye detection function loadCascade() on every new OpenCv object, detection really will be impossible.
I guess without using the specific eye detection function loadCascade() on every new OpenCv object, detection really will be impossible.