Revision history [back]

yes and no ;(

the function itself does not contain any parallel_for_ loops or ocl kernels, but it calls a lot of other functions like filter2D or warpAffine, which are heavily optimized