colorspace.c

mpeg4代码,比较具体

/*****************************************************************************
 *
 *  XVID MPEG-4 VIDEO CODEC
 *  - colorspace conversion module -
 *
 *  Copyright(C) 2002 Peter Ross <[e-mail]>
 *               2002 Michael Militzer <[e-mail]>
 *
 *  This file is part of XviD, a free MPEG-4 video encoder/decoder
 *
 *  XviD 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
 *
 *  Under section 8 of the GNU General Public License, the copyright
 *  holders of XVID explicitly forbid distribution in the following
 *  countries:
 *
 *    - Japan
 *    - United States of America
 *
 *  Linking XviD statically or dynamically with other modules is making a
 *  combined work based on XviD.  Thus, the terms and conditions of the
 *  GNU General Public License cover the whole combination.
 *
 *  As a special exception, the copyright holders of XviD give you
 *  permission to link XviD with independent modules that communicate with
 *  XviD solely through the VFW1.1 and DShow interfaces, regardless of the
 *  license terms of these independent modules, and to copy and distribute
 *  the resulting combined work under terms of your choice, provided that
 *  every copy of the combined work is accompanied by a complete copy of
 *  the source code of XviD (the version of XviD used to produce the
 *  combined work), being distributed under the terms of the GNU General
 *  Public License plus this exception.  An independent module is a module
 *  which is not derived from or based on XviD.
 *
 *  Note that people who make modified versions of XviD are not obligated
 *  to grant this special exception for their modified versions; it is
 *  their choice whether to do so.  The GNU General Public License gives
 *  permission to release a modified version without this exception; this
 *  exception also makes it possible to release a modified version which
 *  carries forward this exception.
 *
 * $Id: colorspace.c,v 1.7 2002/11/26 23:44:10 edgomez Exp $
 *
 ****************************************************************************/

#include <string.h>				/* memcpy */

#include "colorspace.h"
#include "../divx4.h"			/* DEC_PICTURE */

/* function pointers */

/* input */
color_inputFuncPtr rgb555_to_yv12;
color_inputFuncPtr rgb565_to_yv12;
color_inputFuncPtr rgb24_to_yv12;
color_inputFuncPtr rgb32_to_yv12;
color_inputFuncPtr yuv_to_yv12;
color_inputFuncPtr yuyv_to_yv12;
color_inputFuncPtr uyvy_to_yv12;

/* output */
color_outputFuncPtr yv12_to_rgb555;
color_outputFuncPtr yv12_to_rgb565;
color_outputFuncPtr yv12_to_rgb24;
color_outputFuncPtr yv12_to_rgb32;
color_outputFuncPtr yv12_to_yuv;
color_outputFuncPtr yv12_to_yuyv;
color_outputFuncPtr yv12_to_uyvy;

#define MIN(A,B)	((A)<(B)?(A):(B))
#define MAX(A,B)	((A)>(B)?(A):(B))

/*	rgb -> yuv def's

	this following constants are "official spec"
	Video Demystified" (ISBN 1-878707-09-4)

	rgb<->yuv _is_ lossy, since most programs do the conversion differently
		
	SCALEBITS/FIX taken from  ffmpeg
*/

#define Y_R_IN			0.257
#define Y_G_IN			0.504
#define Y_B_IN			0.098
#define Y_ADD_IN		16

#define U_R_IN			0.148
#define U_G_IN			0.291
#define U_B_IN			0.439
#define U_ADD_IN		128

#define V_R_IN			0.439
#define V_G_IN			0.368
#define V_B_IN			0.071
#define V_ADD_IN		128

#define SCALEBITS_IN	8
#define FIX_IN(x)		((uint16_t) ((x) * (1L<<SCALEBITS_IN) + 0.5))


int32_t RGB_Y_tab[256];
int32_t B_U_tab[256];
int32_t G_U_tab[256];
int32_t G_V_tab[256];
int32_t R_V_tab[256];


/* rgb555 -> yuv 4:2:0 planar */
void
rgb555_to_yv12_c(uint8_t * y_out,
				 uint8_t * u_out,
				 uint8_t * v_out,
				 uint8_t * src,
				 int width,
				 int height,
				 int y_stride)
{
	int32_t src_stride = width * 2;
	uint32_t y_dif = y_stride - width;
	uint32_t uv_dif = (y_stride - width) / 2;
	uint32_t x, y;

	if (height < 0) {
		height = -height;
		src += (height - 1) * src_stride;
		src_stride = -src_stride;
	}


	for (y = height / 2; y; y--) {
		/* process one 2x2 block per iteration */
		for (x = 0; x < (uint32_t) width; x += 2) {
			int rgb, r, g, b, r4, g4, b4;

			rgb = *(uint16_t *) (src + x * 2);
			b4 = b = (rgb << 3) & 0xf8;
			g4 = g = (rgb >> 2) & 0xf8;
			r4 = r = (rgb >> 7) & 0xf8;
			y_out[0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + src_stride);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 2) & 0xf8;
			r4 += r = (rgb >> 7) & 0xf8;
			y_out[y_stride] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + 2);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 2) & 0xf8;
			r4 += r = (rgb >> 7) & 0xf8;
			y_out[1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + src_stride + 2);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 2) & 0xf8;
			r4 += r = (rgb >> 7) & 0xf8;
			y_out[y_stride + 1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			*u_out++ =
				(uint8_t) ((-FIX_IN(U_R_IN) * r4 - FIX_IN(U_G_IN) * g4 +
							FIX_IN(U_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				U_ADD_IN;


			*v_out++ =
				(uint8_t) ((FIX_IN(V_R_IN) * r4 - FIX_IN(V_G_IN) * g4 -
							FIX_IN(V_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				V_ADD_IN;

			y_out += 2;
		}
		src += src_stride * 2;
		y_out += y_dif + y_stride;
		u_out += uv_dif;
		v_out += uv_dif;
	}
}



/* rgb565_to_yuv_c
	NOTE:	identical to rgb555 except for shift/mask 
			not tested */

void
rgb565_to_yv12_c(uint8_t * y_out,
				 uint8_t * u_out,
				 uint8_t * v_out,
				 uint8_t * src,
				 int width,
				 int height,
				 int y_stride)
{
	int32_t src_stride = width * 2;

	uint32_t y_dif = y_stride - width;
	uint32_t uv_dif = (y_stride - width) / 2;
	uint32_t x, y;

	if (height < 0) {
		height = -height;
		src += (height - 1) * src_stride;
		src_stride = -src_stride;
	}


	for (y = height / 2; y; y--) {
		/* process one 2x2 block per iteration */
		for (x = 0; x < (uint32_t) width; x += 2) {
			int rgb, r, g, b, r4, g4, b4;

			rgb = *(uint16_t *) (src + x * 2);
			b4 = b = (rgb << 3) & 0xf8;
			g4 = g = (rgb >> 3) & 0xfc;
			r4 = r = (rgb >> 8) & 0xf8;
			y_out[0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + src_stride);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 3) & 0xfc;
			r4 += r = (rgb >> 8) & 0xf8;
			y_out[y_stride] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + 2);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 3) & 0xfc;
			r4 += r = (rgb >> 8) & 0xf8;
			y_out[1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			rgb = *(uint16_t *) (src + x * 2 + src_stride + 2);
			b4 += b = (rgb << 3) & 0xf8;
			g4 += g = (rgb >> 3) & 0xfc;
			r4 += r = (rgb >> 8) & 0xf8;
			y_out[y_stride + 1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			*u_out++ =
				(uint8_t) ((-FIX_IN(U_R_IN) * r4 - FIX_IN(U_G_IN) * g4 +
							FIX_IN(U_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				U_ADD_IN;


			*v_out++ =
				(uint8_t) ((FIX_IN(V_R_IN) * r4 - FIX_IN(V_G_IN) * g4 -
							FIX_IN(V_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				V_ADD_IN;

			y_out += 2;
		}
		src += src_stride * 2;
		y_out += y_dif + y_stride;
		u_out += uv_dif;
		v_out += uv_dif;
	}
}




/*	rgb24 -> yuv 4:2:0 planar 

	NOTE: always flips.
*/

void
rgb24_to_yv12_c(uint8_t * y_out,
				uint8_t * u_out,
				uint8_t * v_out,
				uint8_t * src,
				int width,
				int height,
				int stride)
{
	uint32_t width3 = (width << 1) + width;	/* width * 3 */
	uint32_t src_dif = (width << 3) + width;	/* width3 * 3 */
	uint32_t y_dif = (stride << 1) - width;
	uint32_t uv_dif = (stride - width) >> 1;
	uint32_t x, y;

	src += (height - 2) * width3;


	for (y = height >> 1; y; y--) {
		for (x = width >> 1; x; x--) {
			uint32_t r, g, b, r4, g4, b4;

			b4 = b = src[0];
			g4 = g = src[1];
			r4 = r = src[2];
			y_out[stride + 0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[3]);
			g4 += (g = src[4]);
			r4 += (r = src[5]);
			y_out[stride + 1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[width3 + 0]);
			g4 += (g = src[width3 + 1]);
			r4 += (r = src[width3 + 2]);
			y_out[0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[width3 + 3]);
			g4 += (g = src[width3 + 4]);
			r4 += (r = src[width3 + 5]);
			y_out[1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			*u_out++ =
				(uint8_t) ((-FIX_IN(U_R_IN) * r4 - FIX_IN(U_G_IN) * g4 +
							FIX_IN(U_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				U_ADD_IN;


			*v_out++ =
				(uint8_t) ((FIX_IN(V_R_IN) * r4 - FIX_IN(V_G_IN) * g4 -
							FIX_IN(V_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				V_ADD_IN;


			src += 6;
			y_out += 2;
		}
		src -= src_dif;
		y_out += y_dif;
		u_out += uv_dif;
		v_out += uv_dif;
	}
}


/*	rgb32 -> yuv 4:2:0 planar 

	NOTE: always flips
*/

void
rgb32_to_yv12_c(uint8_t * y_out,
				uint8_t * u_out,
				uint8_t * v_out,
				uint8_t * src,
				int width,
				int height,
				int stride)
{
	uint32_t width4 = (width << 2);	/* width * 4 */
	uint32_t src_dif = 3 * width4;
	uint32_t y_dif = (stride << 1) - width;
	uint32_t uv_dif = (stride - width) >> 1;
	uint32_t x, y;

	src += (height - 2) * width4;

	for (y = height >> 1; y; y--) {
		for (x = width >> 1; x; x--) {
			uint32_t r, g, b, r4, g4, b4;

			b4 = b = src[0];
			g4 = g = src[1];
			r4 = r = src[2];
			y_out[stride + 0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[4]);
			g4 += (g = src[5]);
			r4 += (r = src[6]);
			y_out[stride + 1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[width4 + 0]);
			g4 += (g = src[width4 + 1]);
			r4 += (r = src[width4 + 2]);

			y_out[0] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			b4 += (b = src[width4 + 4]);
			g4 += (g = src[width4 + 5]);
			r4 += (r = src[width4 + 6]);
			y_out[1] =
				(uint8_t) ((FIX_IN(Y_R_IN) * r + FIX_IN(Y_G_IN) * g +
							FIX_IN(Y_B_IN) * b) >> SCALEBITS_IN) + Y_ADD_IN;

			*u_out++ =
				(uint8_t) ((-FIX_IN(U_R_IN) * r4 - FIX_IN(U_G_IN) * g4 +
							FIX_IN(U_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				U_ADD_IN;

			*v_out++ =
				(uint8_t) ((FIX_IN(V_R_IN) * r4 - FIX_IN(V_G_IN) * g4 -
							FIX_IN(V_B_IN) * b4) >> (SCALEBITS_IN + 2)) +
				V_ADD_IN;

			src += 8;
			y_out += 2;
		}
		src -= src_dif;
		y_out += y_dif;
		u_out += uv_dif;
		v_out += uv_dif;
	}
}

/*	yuv planar -> yuv 4:2:0 planar
   
	NOTE: does not flip */

void
yuv_to_yv12_c(uint8_t * y_out,
			  uint8_t * u_out,
			  uint8_t * v_out,
			  uint8_t * src,
			  int width,
			  int height,
			  int stride)
{
	uint32_t stride2 = stride >> 1;
	uint32_t width2 = width >> 1;
	uint32_t y;

	for (y = height; y; y--) {
		memcpy(y_out, src, width);
		src += width;
		y_out += stride;
	}

	for (y = height >> 1; y; y--) {
		memcpy(u_out, src, width2);
		src += width2;
		u_out += stride2;
	}

	for (y = height >> 1; y; y--) {
		memcpy(v_out, src, width2);
		src += width2;
		v_out += stride2;
	}
}



/* yuyv (yuv2) packed -> yuv 4:2:0 planar
   
   NOTE: does not flip */

void
yuyv_to_yv12_c(uint8_t * y_out,
			   uint8_t * u_out,
			   uint8_t * v_out,
			   uint8_t * src,
			   int width,
			   int height,
			   int stride)
{
	uint32_t width2 = width + width;
	uint32_t y_dif = stride - width;
	uint32_t uv_dif = y_dif >> 1;
	uint32_t x, y;

	for (y = height >> 1; y; y--) {

		for (x = width >> 1; x; x--) {
			*y_out++ = *src++;
			/**u_out++ = *src++; */
			*u_out++ = (*(src + width2) + *src) >> 1;
			src++;
			*y_out++ = *src++;
			/**v_out++ = *src++; */
			*v_out++ = (*(src + width2) + *src) >> 1;
			src++;

		}

		y_out += y_dif;
		u_out += uv_dif;
		v_out += uv_dif;

		for (x = width >> 1; x; x--) {
			*y_out++ = *src++;
			src++;
			*y_out++ = *src++;
			src++;
		}

		y_out += y_dif;

	}

}



/* uyvy packed -> yuv 4:2:0 planar
   
   NOTE: does not flip */


void
uyvy_to_yv12_c(uint8_t * y_out,
			   uint8_t * u_out,
			   uint8_t * v_out,
			   uint8_t * src,
			   int width,
			   int height,
			   int stride)
{
	uint32_t width2 = width + width;
	uint32_t y_dif = stride - width;
	uint32_t uv_dif = y_dif >> 1;
	uint32_t x, y;

	for (y = height >> 1; y; y--) {

		for (x = width >> 1; x; x--) {
			*u_out++ = *src++;
			/* *u_out++ = (*(src+width2) + *src++) >> 1; */
			*y_out++ = *src++;
			/**v_out++ = *src++; */
			*v_out++ = (*(src + width2) + *src) >> 1;
			src++;
			*y_out++ = *src++;
		}

		y_out += y_dif;
		u_out += uv_dif;;
		v_out += uv_dif;;

		for (x = width >> 1; x; x--) {
			src++;
			*y_out++ = *src++;
			src++;
			*y_out++ = *src++;
		}

		y_out += y_dif;
	}
}


/*	yuv -> rgb def's */

#define RGB_Y_OUT		1.164
#define B_U_OUT			2.018
#define Y_ADD_OUT		16

#define G_U_OUT			0.391
#define G_V_OUT			0.813
#define U_ADD_OUT		128

#define R_V_OUT			1.596
#define V_ADD_OUT		128


#define SCALEBITS_OUT	13