Bilateral Filter (3D)

Fast 3D recursive bilateral filter implementation based on the one proposed by Qingxiong Yang [17]. More...

Detailed Description

Fast 3D recursive bilateral filter implementation based on the one proposed by Qingxiong Yang [17].

• Possible problems with very large images or images of type UINT32 and bigger

Author

Theodore Chabardes

Collaboration diagram for Bilateral Filter (3D):

Functions
template<class T >
RES_T	recursiveBilateralFilter (const Image< T > &imIn, float sigmaW, float sigmaR, Image< T > &imOut)
	recursiveBilateralFilter 3D Bilateral Filter

Function Documentation

recursiveBilateralFilter()

template<class T >

RES_T recursiveBilateralFilter	(	const Image< T > &	imIn,
		float	sigmaW,
		float	sigmaR,
		Image< T > &	imOut
	)

recursiveBilateralFilter 3D Bilateral Filter

Parameters

[in]	imIn	: input Image
[in]	sigmaW	: spatial standard deviation of kernel
[in]	sigmaR	: range standard deviation of the kernel
[out]	imOut	: output Image

sigmaW and sigmaR heavily depend on the image. A simple rule of thumb to choose initial values to try for sigmaW and sigmaR is :

• sigmaW : smaller than the mean radius of objects to be preserved (e.g. 10 pixels);
• sigmaR : smaller than the minimal gradient to be preserved (e.g. 10 for 8 bits images).