I have a (3D) Histogram which I like to apply Gaussian smoothing on:
cv::MatND Hist;
In the 1D and 2D cases I blur it via:
cv::GaussianBlur(Hist, Hist, cv::Size(1,3), 1.0);// 1D case
cv::GaussianBlur(Hist, Hist, cv::Size(3,3), 1.0);// 2D case
But I struggle to apply Gaussian blurring in the 3D case.
Has anyone got an idea how to attempt this?
Try use separable kernels like shown here: http://www.programming-techniques.com/2013/03/gaussian-blurring-using-separable.html
来源:https://stackoverflow.com/questions/18020438/opencv-gaussian-blur-smoothing-of-3d-matrix-histogram