1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390
|
/*
This simple example shows how zfp can be used to compress 8-bit color images
stored in the PPM image format. This lossy compressor employs two common image
compression strategies: (1) transformation to the YCoCg color space, which
decorrelates color bands, and (2) chroma subsampling, which reduces spatial
resolution in the Co and Cg chrominance bands. The single command-line argument
selects one of two compression modes: if a positive rate (in bits/pixel) is
specified, fixed-rate mode is selected; a negative integer argument, -p, sets
the precision to p in fixed-precision mode. Rate allocation in fixed-rate mode
assigns more bits to luma than to chroma components due to the relatively higher
information content in luma after chroma subsampling.
The YCoCg transform employed here has been adapted to avoid range expansion and
potential overflow. Chroma subsampling is achieved by performing zfp's forward
decorrelating transform and then zeroing all but the four lowest-sequency
coefficients, effectively reducing each chroma block to a bilinear approximation.
Because only four chroma coefficients per 4x4 pixel block are retained, an
alternative to zeroing and then encoding the remaining twelve zero-valued
coefficients is to treat the chroma block as being one-dimensional, with only
four values, and then compressing it using zfp's 1D codec. The dimensionality
of chroma blocks (1 or 2) is specified at compile time via the PPM_CHROMA macro.
NOTE: To keep this example simple, only images whose dimensions are multiples
of four are supported.
*/
#ifdef PPM_CHROMA
#if PPM_CHROMA != 1 && PPM_CHROMA != 2
#error "compile with PPM_CHROMA=1 or PPM_CHROMA=2"
#endif
#else
/* default */
#define PPM_CHROMA 2
#endif
#include <limits.h>
#include <math.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "zfp.h"
/* clamp values to 31-bit range */
static void
clamp(int32* block, uint n)
{
uint i;
for (i = 0; i < n; i++) {
if (block[i] < 1 - (1 << 30))
block[i] = 1 - (1 << 30);
if (block[i] > (1 << 30) - 1)
block[i] = (1 << 30) - 1;
}
}
/* convert 2D block from RGB to YCoCg color space */
static void
rgb2ycocg(int32 ycocg[3][16], /*const*/ int32 rgb[3][16])
{
uint i;
for (i = 0; i < 16; i++) {
int32 r, g, b;
int32 y, co, cg, t;
/* fetch RGB values */
r = rgb[0][i];
g = rgb[1][i];
b = rgb[2][i];
/* perform range-preserving YCoCg forward transform */
co = (r - b) >> 1;
t = b + co;
cg = (g - t) >> 1;
y = t + cg;
/* store YCoCg values */
ycocg[0][i] = y;
ycocg[1][i] = co;
ycocg[2][i] = cg;
}
}
/* convert 2D block from YCoCg to RGB color space */
static void
ycocg2rgb(int32 rgb[3][16], /*const*/ int32 ycocg[3][16])
{
uint i;
for (i = 0; i < 16; i++) {
int32 r, g, b;
int32 y, co, cg, t;
/* fetch YCoCg values */
y = ycocg[0][i];
co = ycocg[1][i];
cg = ycocg[2][i];
/* perform range-preserving YCoCg inverse transform */
t = y - cg;
g = (cg << 1) + t;
b = t - co;
r = (co << 1) + b;
/* store RGB values */
rgb[0][i] = r;
rgb[1][i] = g;
rgb[2][i] = b;
}
}
/* perform partial forward decorrelating transform */
static void
fwd_lift(int32* p, uint s)
{
int32 x, y, z, w;
x = *p; p += s;
y = *p; p += s;
z = *p; p += s;
w = *p; p += s;
x += w; x >>= 1; w -= x;
z += y; z >>= 1; y -= z;
x += z; x >>= 1; z -= x;
w += y; w >>= 1; y -= w;
w += y >> 1; y -= w >> 1;
p -= s; *p = w;
p -= s; *p = z;
p -= s; *p = y;
p -= s; *p = x;
}
/* perform partial inverse decorrelating transform */
static void
inv_lift(int32* p, uint s)
{
int32 x, y, z, w;
x = *p; p += s;
y = *p; p += s;
z = *p; p += s;
w = *p; p += s;
y += w >> 1; w -= y >> 1;
y += w; w <<= 1; w -= y;
z += x; x <<= 1; x -= z;
y += z; z <<= 1; z -= y;
w += x; x <<= 1; x -= w;
p -= s; *p = w;
p -= s; *p = z;
p -= s; *p = y;
p -= s; *p = x;
}
/* perform chroma subsampling by discarding high-frequency components */
static void
chroma_downsample(int32* block)
{
uint i, j;
/* perform forward decorrelating transform */
for (j = 0; j < 4; j++)
fwd_lift(block + 4 * j, 1);
for (i = 0; i < 4; i++)
fwd_lift(block + 1 * i, 4);
#if PPM_CHROMA == 1
/* keep only the four lowest-sequency coefficients */
block[2] = block[4];
block[3] = block[5];
for (i = 4; i < 16; i++)
block[i] = 0;
/* reconstruct as 1D block */
inv_lift(block, 1);
/* clamp values to 31 bits to avoid overflow */
clamp(block, 4);
#else
/* zero out all but four lowest-sequency coefficients */
for (j = 0; j < 4; j++)
for (i = 0; i < 4; i++)
if (i >= 2 || j >= 2)
block[i + 4 * j] = 0;
/* perform inverse decorrelating transform */
for (i = 0; i < 4; i++)
inv_lift(block + 1 * i, 4);
for (j = 0; j < 4; j++)
inv_lift(block + 4 * j, 1);
/* clamp values to 31 bits to avoid overflow */
clamp(block, 16);
#endif
}
/* reconstruct 2D chroma block */
static void
chroma_upsample(int32* block)
{
#if PPM_CHROMA == 1
uint i, j;
/* obtain 1D block coefficients */
fwd_lift(block, 1);
/* reorganize and initialize remaining 2D block coefficients */
block[4] = block[2];
block[5] = block[3];
block[2] = 0;
block[3] = 0;
for (i = 6; i < 16; i++)
block[i] = 0;
/* perform inverse decorrelating transform */
for (i = 0; i < 4; i++)
inv_lift(block + 1 * i, 4);
for (j = 0; j < 4; j++)
inv_lift(block + 4 * j, 1);
/* clamp values to 31 bits to avoid overflow */
clamp(block, 16);
#else
/* clamp values to 31 bits to avoid overflow */
clamp(block, 16);
#endif
}
int main(int argc, char* argv[])
{
double rate = 0;
uint nx, ny;
uint x, y;
uint k;
char line[0x100];
uchar* image;
zfp_field* field;
zfp_stream* zfp[3];
bitstream* stream;
void* buffer;
size_t bytes;
size_t size;
switch (argc) {
case 2:
if (sscanf(argv[1], "%lf", &rate) != 1)
goto usage;
break;
default:
usage:
fprintf(stderr, "Usage: ppm <rate|-precision> <input.ppm >output.ppm\n");
return EXIT_FAILURE;
}
/* read ppm header */
if (!fgets(line, sizeof(line), stdin) || strcmp(line, "P6\n") ||
!fgets(line, sizeof(line), stdin) || sscanf(line, "%u%u", &nx, &ny) != 2 ||
!fgets(line, sizeof(line), stdin) || strcmp(line, "255\n")) {
fprintf(stderr, "error opening image\n");
return EXIT_FAILURE;
}
if ((nx & 3u) || (ny & 3u)) {
fprintf(stderr, "image dimensions must be multiples of four\n");
return EXIT_FAILURE;
}
/* read image data */
image = malloc(3 * nx * ny);
if (!image) {
fprintf(stderr, "error allocating memory\n");
return EXIT_FAILURE;
}
if (fread(image, sizeof(*image), 3 * nx * ny, stdin) != 3 * nx * ny) {
fprintf(stderr, "error reading image\n");
return EXIT_FAILURE;
}
/* initialize compressed streams */
for (k = 0; k < 3; k++)
zfp[k] = zfp_stream_open(NULL);
if (rate < 0) {
/* use fixed-precision mode */
for (k = 0; k < 3; k++)
zfp_stream_set_precision(zfp[k], (uint)floor(0.5 - rate));
}
else {
/* assign higher rate to luminance than to chrominance components */
#if PPM_CHROMA == 1
double chroma_rate = floor(8 * rate / 3 + 0.5) / 4;
double luma_rate = rate - chroma_rate / 2;
zfp_stream_set_rate(zfp[0], luma_rate, zfp_type_int32, 2, zfp_false);
zfp_stream_set_rate(zfp[1], chroma_rate, zfp_type_int32, 1, zfp_false);
zfp_stream_set_rate(zfp[2], chroma_rate, zfp_type_int32, 1, zfp_false);
#else
double chroma_rate = floor(8 * rate / 3 + 0.5) / 16;
double luma_rate = rate - 2 * chroma_rate;
zfp_stream_set_rate(zfp[0], luma_rate, zfp_type_int32, 2, zfp_false);
zfp_stream_set_rate(zfp[1], chroma_rate, zfp_type_int32, 2, zfp_false);
zfp_stream_set_rate(zfp[2], chroma_rate, zfp_type_int32, 2, zfp_false);
#endif
}
/* determine size of compressed buffer */
bytes = 0;
field = zfp_field_2d(image, zfp_type_int32, nx, ny);
for (k = 0; k < 3; k++)
bytes += zfp_stream_maximum_size(zfp[k], field);
zfp_field_free(field);
/* allocate buffer and initialize bit stream */
buffer = malloc(bytes);
if (!buffer) {
fprintf(stderr, "error allocating memory\n");
return EXIT_FAILURE;
}
stream = stream_open(buffer, bytes);
/* the three zfp streams share a single bit stream */
for (k = 0; k < 3; k++)
zfp_stream_set_bit_stream(zfp[k], stream);
/* compress image */
for (y = 0; y < ny; y += 4)
for (x = 0; x < nx; x += 4) {
uchar block[3][16];
int32 rgb[3][16];
int32 ycocg[3][16];
uint i, j, k;
/* fetch R, G, and B blocks */
for (k = 0; k < 3; k++)
for (j = 0; j < 4; j++)
for (i = 0; i < 4; i++)
block[k][i + 4 * j] = image[k + 3 * (x + i + nx * (y + j))];
/* promote to 32-bit integers */
for (k = 0; k < 3; k++)
zfp_promote_uint8_to_int32(rgb[k], block[k], 2);
/* perform color space transform */
rgb2ycocg(ycocg, rgb);
/* chroma subsample the Co and Cg bands */
for (k = 1; k < 3; k++)
chroma_downsample(ycocg[k]);
/* compress the Y, Co, and Cg blocks */
#if PPM_CHROMA == 1
zfp_encode_block_int32_2(zfp[0], ycocg[0]);
zfp_encode_block_int32_1(zfp[1], ycocg[1]);
zfp_encode_block_int32_1(zfp[2], ycocg[2]);
#else
for (k = 0; k < 3; k++)
zfp_encode_block_int32_2(zfp[k], ycocg[k]);
#endif
}
zfp_stream_flush(zfp[0]);
size = zfp_stream_compressed_size(zfp[0]);
fprintf(stderr, "%u compressed bytes (%.2f bits/pixel)\n", (uint)size, (double)size * CHAR_BIT / (nx * ny));
/* decompress image */
zfp_stream_rewind(zfp[0]);
for (y = 0; y < ny; y += 4)
for (x = 0; x < nx; x += 4) {
uchar block[3][16];
int32 rgb[3][16];
int32 ycocg[3][16];
uint i, j, k;
/* decompress the Y, Co, and Cg blocks */
#if PPM_CHROMA == 1
zfp_decode_block_int32_2(zfp[0], ycocg[0]);
zfp_decode_block_int32_1(zfp[1], ycocg[1]);
zfp_decode_block_int32_1(zfp[2], ycocg[2]);
#else
for (k = 0; k < 3; k++)
zfp_decode_block_int32_2(zfp[k], ycocg[k]);
#endif
/* reconstruct Co and Cg chroma bands */
for (k = 1; k < 3; k++)
chroma_upsample(ycocg[k]);
/* perform color space transform */
ycocg2rgb(rgb, ycocg);
/* demote to 8-bit integers */
for (k = 0; k < 3; k++)
zfp_demote_int32_to_uint8(block[k], rgb[k], 2);
/* store R, G, and B blocks */
for (k = 0; k < 3; k++)
for (j = 0; j < 4; j++)
for (i = 0; i < 4; i++)
image[k + 3 * (x + i + nx * (y + j))] = block[k][i + 4 * j];
}
/* clean up */
for (k = 0; k < 3; k++)
zfp_stream_close(zfp[k]);
stream_close(stream);
free(buffer);
/* output reconstructed image */
printf("P6\n");
printf("%u %u\n", nx, ny);
printf("255\n");
if (fwrite(image, sizeof(*image), 3 * nx * ny, stdout) != 3 * nx * ny) {
fprintf(stderr, "error writing image\n");
return EXIT_FAILURE;
}
free(image);
return 0;
}
|