This statement was keeping HAAR cascades from leveraging opencl on nvidia devices. "localSize" on the featureEvaluator remains Size(0, 0) which sets the bool "use_ocl" to false. Adding this allows NVidia GPUs to leverage opencl HAAR Cascades

To avoid problem with reduced stack size of inner threads.

ENV{LIBRARY_PATH} is enough

Use clFinish to gurantee commands completed, instead of waiting for events.
Signed-off-by: Wu Zhiwen <zhiwen.wu@intel.com>