meanshift

The C++ interface for meanshift. More...

array	meanshift (const array &image, const float space, const float color, const int iter=5)
	Meanshift filtering. More...

Detailed Description

The C++ interface for meanshift.

A meanshift filter is an edge-preserving smoothing filter commonly used in object tracking and image segmentation.

This filter replaces each pixel in the image with the mean of the values within a given given color and spatial radius. The meanshift filter is an iterative algorithm that continues until a maxium number of iterations is met or until the value of the means no longer changes.

Function Documentation

array af::meanshift	(	const array &	image,
		const float	space,
		const float	color,
		const int	iter = 5
	)

Meanshift filtering.

// meanshift filter an image with color values [0-1]
float sigma_space = 3.f;
float sigma_color = 0.2f;
int   iterations  = 5;
array img_out = meanshift(img_in, sigma_space, sigma_color, iterations);

// meanshift filter an image with color values [0-255]
float sigma_space = 3.f;
float sigma_color = 0.2f * 255.f;
int   iterations  = 5;
array img_out = meanshift(img_in, sigma_space, sigma_color, iterations);

Parameters

[in]	image	input image
[in]	space	radius in pixels, area for filtering. note: internally space *= 1.5
[in]	color	difference in color allowed, range depends on values in image
[in]	iter	maximum number of iterations for convergence (default=5, max=100)

Note

color range depends on values in the image if image is [0-1], scale color from [0-1]; if image is [0-255], scale color from [0-255];