public BilateralFiltering(float[,] image, float sigmaD, float sigmaR)
{
this.image = image;
int sigmaMax = (int)FastMath.Max(sigmaD, sigmaR);
this.kernelRadius = (int)FastMath.Ceiling(2 * sigmaMax);
float twoSigmaRSquared = 2 * sigmaR * sigmaR;
int kernelSize = this.kernelRadius * 2 + 1;
this.kernelD = new float[kernelSize, kernelSize];
int center = (kernelSize - 1) / 2;
for (int x = -center; x < -center + kernelSize; x++)
{
for (int y = -center; y < -center + kernelSize; y++)
{
this.kernelD[x + center, y + center] = this.gauss(sigmaD, x, y);
}
}
this.gaussSimilarity = new float[256];
for (int i = 0; i < this.gaussSimilarity.Length; i++)
{
this.gaussSimilarity[i] = FastMath.Exp((float)-((i) / twoSigmaRSquared));
}
//rimage = cvCloneImage(image);
}
public BilateralFiltering(float[,] image, float sigmaD, float sigmaR)
{
this.image = image;
var sigmaMax = (int)FastMath.Max(sigmaD, sigmaR);
kernelRadius = (int)FastMath.Ceiling(2 * sigmaMax);
var twoSigmaRSquared = 2 * sigmaR * sigmaR;
var kernelSize = kernelRadius * 2 + 1;
kernelD = new float[kernelSize, kernelSize];
var center = (kernelSize - 1) / 2;
for (var x = -center; x < -center + kernelSize; x++)
{
for (var y = -center; y < -center + kernelSize; y++)
{
kernelD[x + center, y + center] = gauss(sigmaD, x, y);
}
}
gaussSimilarity = new float[256];
for (var i = 0; i < gaussSimilarity.Length; i++)
{
gaussSimilarity[i] = FastMath.Exp(-(i / twoSigmaRSquared));
}
//rimage = cvCloneImage(image);
}
private float gauss(float sigma, int x, int y)
{
return(FastMath.Exp(-((x * x + y * y) / (2 * sigma * sigma))));
}