Learning OpenCV, 2nd Edition

being n

2

in an n-by-n kernel size to being 2n for each pixel. The OpenCV implementation of Gaussian

smoothing also provides even higher performance for several common kernels. 3-by-3, 5-by-5, and 7-by-7

with the “standard” sigma (i.e., sigmaX = 0.0) give better performance than other kernels. Gaussian

blur supports single- or three-channel images in either 8-bit or 32-bit floating-point formats, and it can be

done in place. Results of Gaussian blurring are shown in Figure 5-10.

Bilateral Filter

Figure 5-12: Results of bilateral smoothing

void cv::bilateralFilter(

cv::InputArray src, // Input Image

cv::OutputArray dst, // Result image

int d, // Pixel neighborhood size (max distance)

double sigmaColor, // Width parameter for color weighting function

double sigmaSpace, // Width parameter for spatial weighting function

int borderType = cv::BORDER_DEFAULT // Border extrapolation to use

);

The fifth and final form of smoothing supported by OpenCV is called bilateral filtering [Tomasi98], an

example of which is shown in Figure 5-12. Bilateral filtering is one operation from a somewhat larger class

of image analysis operators known as edge-preserving smoothing. Bilateral filtering is most easily

understood when contrasted to Gaussian smoothing. A typical motivation for Gaussian smoothing is that

pixels in a real image should vary ...