quickshift.c File Reference

Quick shift image segmentation - Definition. More...

#include "quickshift.h"
#include "mathop.h"
#include <string.h>
#include <math.h>
#include <stdio.h>

Functions
vl_qs_type	vl_quickshift_distance (vl_qs_type const *I, int N1, int N2, int K, int i1, int i2, int j1, int j2)
	Computes the accumulated channel L2 distance between i,j + the distance between i,j.
vl_qs_type	vl_quickshift_inner (vl_qs_type const *I, int N1, int N2, int K, int i1, int i2, int j1, int j2)
	Computes the accumulated channel inner product between i,j + the distance between i,j.
VlQS *	vl_quickshift_new (vl_qs_type const *image, int height, int width, int channels)
	Create a quick shift object.
void	vl_quickshift_process (VlQS *q)
	Create a quick shift objet.
void	vl_quickshift_delete (VlQS *q)
	Delete quick shift object.

Detailed Description

For internal use only.

Author:: Brian Fulkerson; Andrea Vedaldi

Definition in file quickshift.c.

Function Documentation

void vl_quickshift_delete ( VlQS * q )

Parameters:

q

quick shift object.

Definition at line 448 of file quickshift.c.

References vl_free().

vl_qs_type vl_quickshift_distance	(	vl_qs_type const *	I,
		int	N1,
		int	N2,
		int	K,
		int	i1,
		int	i2,
		int	j1,
		int	j2
	)			`[inline]`

For internal use only.

Parameters:

	I	input image buffer
	N1	size of the first dimension of the image
	N2	size of the second dimension of the image
	K	number of channels
	i1	first dimension index of the first pixel to compare
	i2	second dimension of the first pixel
	j1	index of the second pixel to compare
	j2	second dimension of the second pixel

Takes the L2 distance between the values in I at pixel i and j, accumulating along K channels and adding in the distance between i,j in the image.

Returns:: the distance as described above

Definition at line 151 of file quickshift.c.

Referenced by vl_quickshift_process().

vl_qs_type vl_quickshift_inner	(	vl_qs_type const *	I,
		int	N1,
		int	N2,
		int	K,
		int	i1,
		int	i2,
		int	j1,
		int	j2
	)			`[inline]`

For internal use only.

Parameters:

	I	input image buffer
	N1	size of the first dimension of the image
	N2	size of the second dimension of the image
	K	number of channels
	i1	first dimension index of the first pixel to compare
	i2	second dimension of the first pixel
	j1	index of the second pixel to compare
	j2	second dimension of the second pixel

Takes the channel-wise inner product between the values in I at pixel i and j, accumulating along K channels and adding in the inner product between i,j in the image.