b9d58ca88b
git-svn-id: https://svn.code.sf.net/p/gphoto/code/trunk/libgphoto2@5969 67ed7778-7388-44ab-90cf-0a291f65f57c
111 lines
5.7 KiB
C
111 lines
5.7 KiB
C
/**********************************************************************
|
|
* Minolta Dimage V digital camera communication library *
|
|
* Copyright © 2000,2001 Gus Hartmann *
|
|
* *
|
|
* This program is free software; you can redistribute it and/or *
|
|
* modify it under the terms of the GNU General Public License as *
|
|
* published by the Free Software Foundation; either version 2 of *
|
|
* the License, or (at your option) any later version. *
|
|
* *
|
|
* This program is distributed in the hope that it will be useful, *
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of *
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
|
|
* GNU General Public License for more details. *
|
|
* *
|
|
* You should have received a copy of the GNU General Public *
|
|
* License along with this program; if not, write to the Free *
|
|
* Software Foundation, Inc., 59 Temple Place, Suite 330, *
|
|
* Boston, MA 02111-1307 USA *
|
|
* *
|
|
**********************************************************************/
|
|
|
|
/* $Id$ */
|
|
|
|
#include "config.h"
|
|
|
|
#include "dimagev.h"
|
|
|
|
#define GP_MODULE "dimagev"
|
|
|
|
/* This function handles the ugliness of converting Y:Cb:Cr data to RGB. The
|
|
return value is a pointer to an array of unsigned chars that has 14413)
|
|
members. The additional thirteen bytes are the PPM "rawbits" header.
|
|
|
|
The Dimage V returns 9600 bytes of data for a thumbnail, in what is referred
|
|
to in the documentation as a "Y:Cb:Cr (4:2:2)" format. This means that the
|
|
data is in a Y:Cr:Cb solor space, and is horizontally compressed so that
|
|
the Cb and Cr values are used for two adjacent bytes. The data essentially
|
|
looks like this:
|
|
|
|
Y(0):Y(1):Cb(0):Cr(0):Y(2):Y(3) ... Y(9599):Y(9600):Cb(9599):Cr(9599)
|
|
|
|
Each of the above is a byte. The image is defined as having 80 pixels of
|
|
width and 60 of height, and this information is not included in the 9600
|
|
bytes of data. Since Y:Cb:Cr data is essentially useless, this function
|
|
converts it into RGB data, with a PPM "rawbits" header containing the
|
|
height and width. The formulae to convert Y:Cb:Cr to RGB are below:
|
|
|
|
B(0,0) = (Cb - 128) * (2 - ( 2 * Cr_Coeff )) + Y(0,0)
|
|
R(0,0) = (Cr - 128) * (2 - ( 2 * Y_Coeff )) + Y(0,0)
|
|
G(0,0) = (Y(0,0) - ( Cr_Coeff * B(0,0) + Y_Coeff * R(0,0)) ) / Cb_Coeff
|
|
|
|
Still with me? Good. The values for the three coefficients all add up to
|
|
one; the values used here are defined in dimagev.h . The code below is
|
|
more complex than this, however; Minolta has not been forthcoming with
|
|
the permitted ranges for the values of Y, Cb, and Cr, so I have had to
|
|
do some guesswork based on the math. I don't know much about colorspaces,
|
|
and I'm not looking to know more. It looks like Cb and Cr have a range
|
|
of (0,128) and Y has a range of (0,255). The Dimage V sometimes returns
|
|
values for Cb and Cr slightly over this range; I crop these to 128. Also,
|
|
sometimes the values returned by the conversion formulae are greater
|
|
than 255, which would overflow the unsigned char used to store them, so
|
|
I crop these to 0. I used to just drop them to 255, but this lead to
|
|
image corruption in dark areas. An afternoon of guesswork later, I tried
|
|
cutting them to 0, and it works well enough for me.
|
|
|
|
I could not have possibly figured this out without the help of these sites:
|
|
http://www.butaman.ne.jp/~tsuruzoh/Computer/Digicams/exif-e.html
|
|
http://www.efg2.com/lab/Graphics/Colors/YUV.htm#411
|
|
http://www.dcs.ed.ac.uk/home/mxr/gfx/2d-hi.html
|
|
*/
|
|
unsigned char *dimagev_ycbcr_to_ppm(unsigned char *ycbcr) {
|
|
unsigned char *rgb_data, *ycrcb_current, *rgb_current;
|
|
int count=0;
|
|
unsigned int magic_r, magic_g, magic_b;
|
|
|
|
if ( ( rgb_data = malloc( 14413 ) ) == NULL ) {
|
|
GP_DEBUG( "dimagev_ycbcr_to_ppm::unable to allocate buffer for Y:Cb:Cr conversion");
|
|
return NULL;
|
|
}
|
|
|
|
ycrcb_current = ycbcr;
|
|
rgb_current = &(rgb_data[13]);
|
|
|
|
/* This is the header for a PPM "rawbits" bitmap of size 80x60. */
|
|
strncpy(rgb_data, "P6\n80 60\n255\n", 13);
|
|
|
|
for ( count = 0 ; count < 9600 ; count+=4, ycrcb_current+=4, rgb_current+=6 ) {
|
|
magic_b = ( ( ycrcb_current[2] > (unsigned char) 128 ? 128 : ycrcb_current[2] ) - 128 ) * ( 2 - ( 2 * CR_COEFF ) ) + ycrcb_current[0];
|
|
rgb_current[2] = (unsigned char) ( magic_b > 255 ? 0 : magic_b );
|
|
magic_r = ( ( ycrcb_current[3] > (unsigned char) 128 ? 128 : ycrcb_current[3] ) - 128 ) * ( 2 - ( 2 * Y_COEFF ) ) + ycrcb_current[0];
|
|
rgb_current[0] = (unsigned char) ( magic_r > 255 ? 0 : magic_r );
|
|
magic_g = (( ycrcb_current[0] - ( CR_COEFF * rgb_current[2] ) ) - ( Y_COEFF * rgb_current[0])) / CB_COEFF ;
|
|
rgb_current[1] = (unsigned char) ( magic_g > 255 ? 0 : magic_g );
|
|
|
|
/* Wipe everything clean. */
|
|
magic_b = magic_r = magic_g = 0;
|
|
|
|
magic_b = ( ( ycrcb_current[2] > (unsigned char) 128 ? 128 : ycrcb_current[2] ) - 128 ) * ( 2 - ( 2 * CR_COEFF ) ) + ycrcb_current[1];
|
|
rgb_current[5] = (unsigned char) ( magic_b > 255 ? 0 : magic_b );
|
|
magic_r = ( ( ycrcb_current[3] > (unsigned char) 128 ? 128 : ycrcb_current[3] ) - 128 ) * ( 2 - ( 2 * Y_COEFF ) ) + ycrcb_current[1];
|
|
rgb_current[3] = (unsigned char) ( magic_r > 255 ? 0 : magic_r );
|
|
magic_g = (( ycrcb_current[1] - ( CR_COEFF * rgb_current[5] ) ) - ( Y_COEFF * rgb_current[3])) / CB_COEFF ;
|
|
rgb_current[4] = (unsigned char) ( magic_g > 255 ? 0 : magic_g );
|
|
|
|
/* Wipe everything clean. */
|
|
magic_b = magic_r = magic_g = 0;
|
|
}
|
|
|
|
return rgb_data;
|
|
}
|