2d.c File Reference

Definitions for matrix filters. More...

#include <jwsc/config.h>
#include <stdlib.h>
#include <inttypes.h>
#include <math.h>
#include <assert.h>
#include "jwsc/base/error.h"
#include "jwsc/prob/pdf.h"
#include "jwsc/matrix/matrix.h"
#include "jwsc/matrix/matrix_math.h"
#include "jwsc/filter/2d.h"

Include dependency graph for 2d.c:

Go to the source code of this file.

Defines
#define	PI   3.14159265358979323846
	Estimate of pi.
Functions
create_sobel_x_filter
Creates a Sobel operator for the x direction.
void	create_sobel_x_filter_f (Matrix_f **sobel_out)
	Creates a single precision Sobel operator for the x direction.
void	create_sobel_x_filter_d (Matrix_d **sobel_out)
	Creates a double precision Sobel operator for the x direction.
create_sobel_y_filter
Creates a Sobel operator for the y direction.
void	create_sobel_y_filter_f (Matrix_f **sobel_out)
	Creates a single precision Sobel operator for the y direction.
void	create_sobel_y_filter_d (Matrix_d **sobel_out)
	Creates a double precision Sobel operator for the y direction.
create_LoG_filter
Creates a Laplacian of Gaussian filter.
Error *	create_LoG_filter_f (Matrix_f **LoG_out, float sigma)
	Creates a single precision Laplacian of Gaussian filter.
create_auto_2d_gaussian_filter
Creates a 2D Gaussian filter automatically sized to sqrt(2)*6*sigma.
Error *	create_auto_2d_gaussian_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma)
	Creates a single precision 2D Gaussian filter automatically sized.
Error *	create_auto_2d_gaussian_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma)
	Creates a double precision 2D Gaussian filter automatically sized.
create_2d_gaussian_filter
Creates a 2D Gaussian filter.
Error *	create_2d_gaussian_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a single precision 2D Gaussian filter.
Error *	create_2d_gaussian_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a double precision 2D Gaussian filter.
create_auto_2d_gaussian_dx_filter
Creates a 2D Gaussian filter with a partial derivative along columns automatically sized to sqrt(2)*6*sigma.
Error *	create_auto_2d_gaussian_dx_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma)
	Creates a single precision 2D Gaussian filter with a partial derivative in the x-direction (column) automatically sized.
Error *	create_auto_2d_gaussian_dx_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma)
	Creates a double precision 2D Gaussian filter with a partial derivative in the x-direction (column) automatically sized.
create_2d_gaussian_dx_filter
Creates a 2D Gaussian filter with a partial derivative along columns.
Error *	create_2d_gaussian_dx_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a single precision 2D Gaussian filter with a partial derivative in the x-direction (column).
Error *	create_2d_gaussian_dx_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a double precision 2D Gaussian filter with a partial derivative in the x-direction (column).
create_auto_2d_gaussian_dy_filter
Creates a 2D Gaussian filter with a partial derivative along rows automatically sized to sqrt(2)*6*sigma.
Error *	create_auto_2d_gaussian_dy_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma)
	Creates a single precision 2D Gaussian filter with a partial derivative in the y-direction (row) automatically sized.
Error *	create_auto_2d_gaussian_dy_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma)
	Creates a double precision 2D Gaussian filter with a partial derivative in the y-direction (row) automatically sized.
create_2d_gaussian_dy_filter
Creates a 2D Gaussian filter with a partial derivative along rows.
Error *	create_2d_gaussian_dy_filter_f (Matrix_f **h_out, float row_sigma, float col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a single precision 2D Gaussian filter with a partial derivative in the y direction (row).
Error *	create_2d_gaussian_dy_filter_d (Matrix_d **h_out, double row_sigma, double col_sigma, uint32_t num_rows, uint32_t num_cols)
	Creates a double precision 2D Gaussian filter with a partial derivative in the y direction (row).

---

Detailed Description

Definitions for matrix filters.

Author:

Joseph Schlecht

License:

Creative Commons BY-NC-SA 3.0

Definition in file 2d.c.

---

Define Documentation

#define PI   3.14159265358979323846

Estimate of pi.

Definition at line 61 of file 2d.c.

---

Function Documentation

void create_sobel_x_filter_f

(

Matrix_f **

sobel_out

)

Creates a single precision Sobel operator for the x direction.

The x direction is across matrix columns.

Parameters:

sobel_out

Result parameter. If *sobel_out is NULL, a matrix is allocated; otherwise its space is re-used.

Note:

Does not scale the operator.

Definition at line 79 of file 2d.c.

void create_sobel_x_filter_d

(

Matrix_d **

sobel_out

)

Creates a double precision Sobel operator for the x direction.

The x direction is across matrix columns.

Parameters:

sobel_out

Result parameter. If *sobel_out is NULL, a matrix is allocated; otherwise its space is re-used.

Note:

Does not scale the operator.

Definition at line 102 of file 2d.c.

void create_sobel_y_filter_f

(

Matrix_f **

sobel_out

)

Creates a single precision Sobel operator for the y direction.

The y direction is across matrix rows.

Parameters:

sobel_out

Result parameter. If *sobel_out is NULL, a matrix is allocated; otherwise its space is re-used.

Note:

Does not scale the operator.

Definition at line 137 of file 2d.c.

void create_sobel_y_filter_d

(

Matrix_d **

sobel_out

)

Creates a double precision Sobel operator for the y direction.

The y direction is across matrix rows.

Parameters:

sobel_out

Result parameter. If *sobel_out is NULL, a matrix is allocated; otherwise its space is re-used.

Note:

Does not scale the operator.

Definition at line 160 of file 2d.c.

Error* create_LoG_filter_f	(	Matrix_f **	LoG_out,
		float	sigma
	)

Creates a single precision Laplacian of Gaussian filter.

Parameters:

LoG_out	Result parameter. If *LoG_out is NULL, a matrix is allocated; otherwise its space is re-used.
sigma	Sigma to the Laplacian.

Returns:

On success, NULL is returned. On error, an Error is returned and *LoG_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Does not scale the operator.

The dimensions of the filter will be square and odd.

Definition at line 201 of file 2d.c.

Error* create_auto_2d_gaussian_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma
	)

Creates a single precision 2D Gaussian filter automatically sized.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Does not normalize the sum of the filter to 1.

The dimensions of the filter will be odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 268 of file 2d.c.

Error* create_auto_2d_gaussian_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma
	)

Creates a double precision 2D Gaussian filter automatically sized.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Does not normalize the sum of the filter to 1.

The dimensions of the filter will be odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 300 of file 2d.c.

Error* create_2d_gaussian_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a single precision 2D Gaussian filter.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Does not normalize the sum of the filter to 1.

If integrate_gaussian_pdf_d() function is not available, then the gaussian_pdf_d() is used for each value of the matrix.

Dimensions must be > 0.

Definition at line 348 of file 2d.c.

Error* create_2d_gaussian_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a double precision 2D Gaussian filter.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Does not normalize the sum of the filter to 1.

If integrate_gaussian_pdf_d() function is not available, then the gaussian_pdf_d() is used for each value of the matrix.

Dimensions must be > 0.

Definition at line 447 of file 2d.c.

Error* create_auto_2d_gaussian_dx_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma
	)

Creates a single precision 2D Gaussian filter with a partial derivative in the x-direction (column) automatically sized.

The x direction is across matrix columns.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

The dimensions of the filter will be square and odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 555 of file 2d.c.

Error* create_auto_2d_gaussian_dx_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma
	)

Creates a double precision 2D Gaussian filter with a partial derivative in the x-direction (column) automatically sized.

The x direction is across matrix columns.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

The dimensions of the filter will be square and odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 587 of file 2d.c.

Error* create_2d_gaussian_dx_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a single precision 2D Gaussian filter with a partial derivative in the x-direction (column).

The x direction is across matrix columns.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Dimensions must be > 0.

Definition at line 632 of file 2d.c.

Error* create_2d_gaussian_dx_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a double precision 2D Gaussian filter with a partial derivative in the x-direction (column).

The x direction is across matrix columns.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Dimensions must be > 0.

Definition at line 699 of file 2d.c.

Error* create_auto_2d_gaussian_dy_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma
	)

Creates a single precision 2D Gaussian filter with a partial derivative in the y-direction (row) automatically sized.

The y direction is across matrix rows.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

The dimensions of the filter will be square and odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 779 of file 2d.c.

Error* create_auto_2d_gaussian_dy_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma
	)

Creates a double precision 2D Gaussian filter with a partial derivative in the y-direction (row) automatically sized.

The y direction is across matrix rows.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

The dimensions of the filter will be square and odd and enough to contain at least 6 stdev of the Gaussian.

Definition at line 811 of file 2d.c.

Error* create_2d_gaussian_dy_filter_f	(	Matrix_f **	h_out,
		float	row_sigma,
		float	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a single precision 2D Gaussian filter with a partial derivative in the y direction (row).

The y direction is across matrix rows.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Dimensions must be > 0.

Definition at line 856 of file 2d.c.

Error* create_2d_gaussian_dy_filter_d	(	Matrix_d **	h_out,
		double	row_sigma,
		double	col_sigma,
		uint32_t	num_rows,
		uint32_t	num_cols
	)

Creates a double precision 2D Gaussian filter with a partial derivative in the y direction (row).

The y direction is across matrix rows.

Parameters:

h_out	Result parameter. If *h_out is NULL, a matrix is allocated; otherwise its space is re-used.
row_sigma	Sigma to the Gaussian along rows.
col_sigma	Sigma to the Gaussian along columns.
num_rows	Number of rows.
num_cols	Number of columns.

Returns:

On success, NULL is returned. On error, an Error is returned and *h_out is freed and set to NULL.

• ERROR_INV_ARG sigma is too small.

Note:

Dimensions must be > 0.

Definition at line 923 of file 2d.c.