pgm.h File Reference

Portable graymap format (PGM) parser. More...

#include "generic.h"
#include "mathop.h"
#include <stdio.h>

Data Structures
struct	VlPgmImage
	PGM image meta data. More...

Macros
PGM parser error codes
#define	VL_ERR_PGM_INV_HEAD   101
#define	VL_ERR_PGM_INV_META   102
#define	VL_ERR_PGM_INV_DATA   103
#define	VL_ERR_PGM_IO   104

Functions
Core operations
int	vl_pgm_extract_head (FILE *f, VlPgmImage *im)
	Extract PGM header from stream. More...
int	vl_pgm_extract_data (FILE *f, VlPgmImage const *im, void *data)
	Extract PGM data from stream. More...
int	vl_pgm_insert (FILE *f, VlPgmImage const *im, void const *data)
	Insert a PGM image into a stream. More...
vl_size	vl_pgm_get_npixels (VlPgmImage const *im)
	Get PGM image number of pixels. More...
vl_size	vl_pgm_get_bpp (VlPgmImage const *im)
	Get PGM image bytes per pixel. More...
Helper functions
int	vl_pgm_write (char const *name, vl_uint8 const *data, int width, int height)
	Write bytes to a PGM file. More...
int	vl_pgm_write_f (char const *name, float const *data, int width, int height)
	Write floats to PGM file. More...
int	vl_pgm_read_new (char const *name, VlPgmImage *im, vl_uint8 **data)
	Read a PGM file. More...
int	vl_pgm_read_new_f (char const *name, VlPgmImage *im, float **data)
	Read floats from a PGM file. More...

Detailed Description

This module implements basic input and ouptut of images in PGM format.

Extracting an image encoded in PGM format from an imput file stream involves the following steps:

• use vl_pgm_extract_head to extract the image meta data (size and bit depth);
• allocate a buffer to store the image data;
• use vl_pgm_extract_data to extract the image data to the allocated buffer.

Writing an image in PGM format to an ouptut file stream can be done by using vl_pgm_insert.

To quickly read/write a PGM image from/to a given file, use vl_pgm_read_new() and vl_pgm_write(). To to the same from a buffer in floating point format use vl_pgm_read_new_f() and vl_pgm_write_f().

Author

Andrea Vedaldi

Macro Definition Documentation

VL_ERR_PGM_INV_DATA

#define VL_ERR_PGM_INV_DATA   103

Invalid PGM data section.

VL_ERR_PGM_INV_HEAD

#define VL_ERR_PGM_INV_HEAD   101

Invalid PGM header section.

VL_ERR_PGM_INV_META

#define VL_ERR_PGM_INV_META   102

Invalid PGM meta section.

VL_ERR_PGM_IO

#define VL_ERR_PGM_IO   104

Generic I/O error.

Function Documentation

vl_pgm_extract_data()

int vl_pgm_extract_data	(	FILE *	f,
		VlPgmImage const *	im,
		void *	data
	)

Parameters

f	input file.
im	PGM image descriptor.
data	data buffer to fill.

Returns

error code.

The function extracts from the file f the data section of an image encoded in PGM format. The function fills the buffer data according. The buffer data should be vl_pgm_get_npixels() by vl_pgm_get_bpp() bytes large.

vl_pgm_extract_head()

int vl_pgm_extract_head	(	FILE *	f,
		VlPgmImage *	im
	)

Parameters

f	input file.
im	image structure to fill.

Returns

error code.

The function extracts from the file f the meta-data section of an image encoded in PGM format. The function fills the structure VlPgmImage accordingly.

The error may be either VL_ERR_PGM_INV_HEAD or VL_ERR_PGM_INV_META depending whether the error occurred in decoding the header or meta section of the PGM file.

vl_pgm_get_bpp()

vl_size vl_pgm_get_bpp

(

VlPgmImage const *

im

)

Parameters

im	PGM image descriptor.

Returns

number of bytes per pixel.

The function returns the number of bytes for each pixel of the PGM image im.

vl_pgm_get_npixels()

vl_size vl_pgm_get_npixels

(

VlPgmImage const *

im

)

Parameters

im	PGM image descriptor.

Returns

number of pixels of the image.

The functions returns the number of pixels of the PGM image im.

To calculate the image data size in bytes, this value must be multiplied by the number of byte per pixels (see vl_pgm_get_bpp()).

vl_pgm_insert()

int vl_pgm_insert	(	FILE *	f,
		VlPgmImage const *	im,
		void const *	data
	)

Parameters

f	output file.
im	PGM image meta-data.
data	image data.

Returns

error code.

vl_pgm_read_new()

int vl_pgm_read_new	(	char const *	name,
		VlPgmImage *	im,
		vl_uint8 **	data
	)

Parameters

name	file name.
im	a pointer to the PGM image structure to fill.
data	a pointer to the pointer to the allocated buffer.

Returns

error code.

The function reads a PGM image from file name and initializes the structure im and the buffer data accordingly.

The ownership of the buffer data is transfered to the caller. data should be freed by means of vl_free().

Bug:

Only PGM files with 1 BPP are supported.

vl_pgm_read_new_f()

int vl_pgm_read_new_f	(	char const *	name,
		VlPgmImage *	im,
		float **	data
	)

Parameters

name	file name.
im	a pointer to the PGM image structure to fill.
data	a pointer to the pointer to the allocated buffer.

Returns

error code.

The function reads a PGM image from file name and initializes the structure im and the buffer data accordingly. The buffer data is an array of floats in the range [0, 1].

The ownership of the buffer data is transfered to the caller. data should be freed by means of vl_free().

Bug:

Only PGM files with 1 BPP are supported.

vl_pgm_write()

int vl_pgm_write	(	char const *	name,
		vl_uint8 const *	data,
		int	width,
		int	height
	)

Parameters

name	file name.
data	data to write.
width	width of the image.
height	height of the image.

Returns

error code.

The function dumps the image data to the PGM file of the specified name. This is an helper function simplifying the usage of vl_pgm_insert().

vl_pgm_write_f()

int vl_pgm_write_f	(	char const *	name,
		float const *	data,
		int	width,
		int	height
	)

Parameters

name	file name.
data	data to write.
width	width of the image.
height	height of the image.

Returns

error code.

The function dumps the image data to the PGM file of the specified name. The data is re-scaled to fit in the range 0-255. This is an helper function simplifying the usage of vl_pgm_insert().