FFmpeg
 All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Groups Pages
snow_dwt.c
Go to the documentation of this file.
1 /*
2  * Copyright (C) 2004-2010 Michael Niedermayer <michaelni@gmx.at>
3  * Copyright (C) 2008 David Conrad
4  *
5  * This file is part of FFmpeg.
6  *
7  * FFmpeg is free software; you can redistribute it and/or
8  * modify it under the terms of the GNU Lesser General Public
9  * License as published by the Free Software Foundation; either
10  * version 2.1 of the License, or (at your option) any later version.
11  *
12  * FFmpeg is distributed in the hope that it will be useful,
13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
15  * Lesser General Public License for more details.
16  *
17  * You should have received a copy of the GNU Lesser General Public
18  * License along with FFmpeg; if not, write to the Free Software
19  * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
20  */
21 
22 #include "libavutil/attributes.h"
23 #include "libavutil/avassert.h"
24 #include "libavutil/common.h"
25 #include "me_cmp.h"
26 #include "snow_dwt.h"
27 
28 int ff_slice_buffer_init(slice_buffer *buf, int line_count,
29  int max_allocated_lines, int line_width,
30  IDWTELEM *base_buffer)
31 {
32  int i;
33 
34  buf->base_buffer = base_buffer;
35  buf->line_count = line_count;
36  buf->line_width = line_width;
37  buf->data_count = max_allocated_lines;
38  buf->line = av_mallocz_array(line_count, sizeof(IDWTELEM *));
39  if (!buf->line)
40  return AVERROR(ENOMEM);
41  buf->data_stack = av_malloc_array(max_allocated_lines, sizeof(IDWTELEM *));
42  if (!buf->data_stack) {
43  av_freep(&buf->line);
44  return AVERROR(ENOMEM);
45  }
46 
47  for (i = 0; i < max_allocated_lines; i++) {
48  buf->data_stack[i] = av_malloc_array(line_width, sizeof(IDWTELEM));
49  if (!buf->data_stack[i]) {
50  for (i--; i >=0; i--)
51  av_freep(&buf->data_stack[i]);
52  av_freep(&buf->data_stack);
53  av_freep(&buf->line);
54  return AVERROR(ENOMEM);
55  }
56  }
57 
58  buf->data_stack_top = max_allocated_lines - 1;
59  return 0;
60 }
61 
62 IDWTELEM *ff_slice_buffer_load_line(slice_buffer *buf, int line)
63 {
64  IDWTELEM *buffer;
65 
66  av_assert0(buf->data_stack_top >= 0);
67 // av_assert1(!buf->line[line]);
68  if (buf->line[line])
69  return buf->line[line];
70 
71  buffer = buf->data_stack[buf->data_stack_top];
72  buf->data_stack_top--;
73  buf->line[line] = buffer;
74 
75  return buffer;
76 }
77 
78 void ff_slice_buffer_release(slice_buffer *buf, int line)
79 {
80  IDWTELEM *buffer;
81 
82  av_assert1(line >= 0 && line < buf->line_count);
83  av_assert1(buf->line[line]);
84 
85  buffer = buf->line[line];
86  buf->data_stack_top++;
87  buf->data_stack[buf->data_stack_top] = buffer;
88  buf->line[line] = NULL;
89 }
90 
91 void ff_slice_buffer_flush(slice_buffer *buf)
92 {
93  int i;
94 
95  if (!buf->line)
96  return;
97 
98  for (i = 0; i < buf->line_count; i++)
99  if (buf->line[i])
100  ff_slice_buffer_release(buf, i);
101 }
102 
103 void ff_slice_buffer_destroy(slice_buffer *buf)
104 {
105  int i;
106  ff_slice_buffer_flush(buf);
107 
108  if (buf->data_stack)
109  for (i = buf->data_count - 1; i >= 0; i--)
110  av_freep(&buf->data_stack[i]);
111  av_freep(&buf->data_stack);
112  av_freep(&buf->line);
113 }
114 
115 static av_always_inline void lift(DWTELEM *dst, DWTELEM *src, DWTELEM *ref,
116  int dst_step, int src_step, int ref_step,
117  int width, int mul, int add, int shift,
118  int highpass, int inverse)
119 {
120  const int mirror_left = !highpass;
121  const int mirror_right = (width & 1) ^ highpass;
122  const int w = (width >> 1) - 1 + (highpass & width);
123  int i;
124 
125 #define LIFT(src, ref, inv) ((src) + ((inv) ? -(ref) : +(ref)))
126  if (mirror_left) {
127  dst[0] = LIFT(src[0], ((mul * 2 * ref[0] + add) >> shift), inverse);
128  dst += dst_step;
129  src += src_step;
130  }
131 
132  for (i = 0; i < w; i++)
133  dst[i * dst_step] = LIFT(src[i * src_step],
134  ((mul * (ref[i * ref_step] +
135  ref[(i + 1) * ref_step]) +
136  add) >> shift),
137  inverse);
138 
139  if (mirror_right)
140  dst[w * dst_step] = LIFT(src[w * src_step],
141  ((mul * 2 * ref[w * ref_step] + add) >> shift),
142  inverse);
143 }
144 
145 static av_always_inline void liftS(DWTELEM *dst, DWTELEM *src, DWTELEM *ref,
146  int dst_step, int src_step, int ref_step,
147  int width, int mul, int add, int shift,
148  int highpass, int inverse)
149 {
150  const int mirror_left = !highpass;
151  const int mirror_right = (width & 1) ^ highpass;
152  const int w = (width >> 1) - 1 + (highpass & width);
153  int i;
154 
155  av_assert1(shift == 4);
156 #define LIFTS(src, ref, inv) \
157  ((inv) ? (src) + (((ref) + 4 * (src)) >> shift) \
158  : -((-16 * (src) + (ref) + add / \
159  4 + 1 + (5 << 25)) / (5 * 4) - (1 << 23)))
160  if (mirror_left) {
161  dst[0] = LIFTS(src[0], mul * 2 * ref[0] + add, inverse);
162  dst += dst_step;
163  src += src_step;
164  }
165 
166  for (i = 0; i < w; i++)
167  dst[i * dst_step] = LIFTS(src[i * src_step],
168  mul * (ref[i * ref_step] +
169  ref[(i + 1) * ref_step]) + add,
170  inverse);
171 
172  if (mirror_right)
173  dst[w * dst_step] = LIFTS(src[w * src_step],
174  mul * 2 * ref[w * ref_step] + add,
175  inverse);
176 }
177 
178 static void horizontal_decompose53i(DWTELEM *b, DWTELEM *temp, int width)
179 {
180  const int width2 = width >> 1;
181  int x;
182  const int w2 = (width + 1) >> 1;
183 
184  for (x = 0; x < width2; x++) {
185  temp[x] = b[2 * x];
186  temp[x + w2] = b[2 * x + 1];
187  }
188  if (width & 1)
189  temp[x] = b[2 * x];
190  lift(b + w2, temp + w2, temp, 1, 1, 1, width, -1, 0, 1, 1, 0);
191  lift(b, temp, b + w2, 1, 1, 1, width, 1, 2, 2, 0, 0);
192 }
193 
194 static void vertical_decompose53iH0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
195  int width)
196 {
197  int i;
198 
199  for (i = 0; i < width; i++)
200  b1[i] -= (b0[i] + b2[i]) >> 1;
201 }
202 
203 static void vertical_decompose53iL0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
204  int width)
205 {
206  int i;
207 
208  for (i = 0; i < width; i++)
209  b1[i] += (b0[i] + b2[i] + 2) >> 2;
210 }
211 
212 static void spatial_decompose53i(DWTELEM *buffer, DWTELEM *temp,
213  int width, int height, int stride)
214 {
215  int y;
216  DWTELEM *b0 = buffer + avpriv_mirror(-2 - 1, height - 1) * stride;
217  DWTELEM *b1 = buffer + avpriv_mirror(-2, height - 1) * stride;
218 
219  for (y = -2; y < height; y += 2) {
220  DWTELEM *b2 = buffer + avpriv_mirror(y + 1, height - 1) * stride;
221  DWTELEM *b3 = buffer + avpriv_mirror(y + 2, height - 1) * stride;
222 
223  if (y + 1 < (unsigned)height)
224  horizontal_decompose53i(b2, temp, width);
225  if (y + 2 < (unsigned)height)
226  horizontal_decompose53i(b3, temp, width);
227 
228  if (y + 1 < (unsigned)height)
229  vertical_decompose53iH0(b1, b2, b3, width);
230  if (y + 0 < (unsigned)height)
231  vertical_decompose53iL0(b0, b1, b2, width);
232 
233  b0 = b2;
234  b1 = b3;
235  }
236 }
237 
238 static void horizontal_decompose97i(DWTELEM *b, DWTELEM *temp, int width)
239 {
240  const int w2 = (width + 1) >> 1;
241 
242  lift(temp + w2, b + 1, b, 1, 2, 2, width, W_AM, W_AO, W_AS, 1, 1);
243  liftS(temp, b, temp + w2, 1, 2, 1, width, W_BM, W_BO, W_BS, 0, 0);
244  lift(b + w2, temp + w2, temp, 1, 1, 1, width, W_CM, W_CO, W_CS, 1, 0);
245  lift(b, temp, b + w2, 1, 1, 1, width, W_DM, W_DO, W_DS, 0, 0);
246 }
247 
248 static void vertical_decompose97iH0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
249  int width)
250 {
251  int i;
252 
253  for (i = 0; i < width; i++)
254  b1[i] -= (W_AM * (b0[i] + b2[i]) + W_AO) >> W_AS;
255 }
256 
257 static void vertical_decompose97iH1(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
258  int width)
259 {
260  int i;
261 
262  for (i = 0; i < width; i++)
263  b1[i] += (W_CM * (b0[i] + b2[i]) + W_CO) >> W_CS;
264 }
265 
266 static void vertical_decompose97iL0(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
267  int width)
268 {
269  int i;
270 
271  for (i = 0; i < width; i++)
272  b1[i] = (16 * 4 * b1[i] - 4 * (b0[i] + b2[i]) + W_BO * 5 + (5 << 27)) /
273  (5 * 16) - (1 << 23);
274 }
275 
276 static void vertical_decompose97iL1(DWTELEM *b0, DWTELEM *b1, DWTELEM *b2,
277  int width)
278 {
279  int i;
280 
281  for (i = 0; i < width; i++)
282  b1[i] += (W_DM * (b0[i] + b2[i]) + W_DO) >> W_DS;
283 }
284 
285 static void spatial_decompose97i(DWTELEM *buffer, DWTELEM *temp,
286  int width, int height, int stride)
287 {
288  int y;
289  DWTELEM *b0 = buffer + avpriv_mirror(-4 - 1, height - 1) * stride;
290  DWTELEM *b1 = buffer + avpriv_mirror(-4, height - 1) * stride;
291  DWTELEM *b2 = buffer + avpriv_mirror(-4 + 1, height - 1) * stride;
292  DWTELEM *b3 = buffer + avpriv_mirror(-4 + 2, height - 1) * stride;
293 
294  for (y = -4; y < height; y += 2) {
295  DWTELEM *b4 = buffer + avpriv_mirror(y + 3, height - 1) * stride;
296  DWTELEM *b5 = buffer + avpriv_mirror(y + 4, height - 1) * stride;
297 
298  if (y + 3 < (unsigned)height)
299  horizontal_decompose97i(b4, temp, width);
300  if (y + 4 < (unsigned)height)
301  horizontal_decompose97i(b5, temp, width);
302 
303  if (y + 3 < (unsigned)height)
304  vertical_decompose97iH0(b3, b4, b5, width);
305  if (y + 2 < (unsigned)height)
306  vertical_decompose97iL0(b2, b3, b4, width);
307  if (y + 1 < (unsigned)height)
308  vertical_decompose97iH1(b1, b2, b3, width);
309  if (y + 0 < (unsigned)height)
310  vertical_decompose97iL1(b0, b1, b2, width);
311 
312  b0 = b2;
313  b1 = b3;
314  b2 = b4;
315  b3 = b5;
316  }
317 }
318 
319 void ff_spatial_dwt(DWTELEM *buffer, DWTELEM *temp, int width, int height,
320  int stride, int type, int decomposition_count)
321 {
322  int level;
323 
324  for (level = 0; level < decomposition_count; level++) {
325  switch (type) {
326  case DWT_97:
327  spatial_decompose97i(buffer, temp,
328  width >> level, height >> level,
329  stride << level);
330  break;
331  case DWT_53:
332  spatial_decompose53i(buffer, temp,
333  width >> level, height >> level,
334  stride << level);
335  break;
336  }
337  }
338 }
339 
340 static void horizontal_compose53i(IDWTELEM *b, IDWTELEM *temp, int width)
341 {
342  const int width2 = width >> 1;
343  const int w2 = (width + 1) >> 1;
344  int x;
345 
346  for (x = 0; x < width2; x++) {
347  temp[2 * x] = b[x];
348  temp[2 * x + 1] = b[x + w2];
349  }
350  if (width & 1)
351  temp[2 * x] = b[x];
352 
353  b[0] = temp[0] - ((temp[1] + 1) >> 1);
354  for (x = 2; x < width - 1; x += 2) {
355  b[x] = temp[x] - ((temp[x - 1] + temp[x + 1] + 2) >> 2);
356  b[x - 1] = temp[x - 1] + ((b[x - 2] + b[x] + 1) >> 1);
357  }
358  if (width & 1) {
359  b[x] = temp[x] - ((temp[x - 1] + 1) >> 1);
360  b[x - 1] = temp[x - 1] + ((b[x - 2] + b[x] + 1) >> 1);
361  } else
362  b[x - 1] = temp[x - 1] + b[x - 2];
363 }
364 
365 static void vertical_compose53iH0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
366  int width)
367 {
368  int i;
369 
370  for (i = 0; i < width; i++)
371  b1[i] += (b0[i] + b2[i]) >> 1;
372 }
373 
374 static void vertical_compose53iL0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
375  int width)
376 {
377  int i;
378 
379  for (i = 0; i < width; i++)
380  b1[i] -= (b0[i] + b2[i] + 2) >> 2;
381 }
382 
383 static void spatial_compose53i_buffered_init(DWTCompose *cs, slice_buffer *sb,
384  int height, int stride_line)
385 {
386  cs->b0 = slice_buffer_get_line(sb,
387  avpriv_mirror(-1 - 1, height - 1) * stride_line);
388  cs->b1 = slice_buffer_get_line(sb, avpriv_mirror(-1, height - 1) * stride_line);
389  cs->y = -1;
390 }
391 
392 static void spatial_compose53i_init(DWTCompose *cs, IDWTELEM *buffer,
393  int height, int stride)
394 {
395  cs->b0 = buffer + avpriv_mirror(-1 - 1, height - 1) * stride;
396  cs->b1 = buffer + avpriv_mirror(-1, height - 1) * stride;
397  cs->y = -1;
398 }
399 
400 static void spatial_compose53i_dy_buffered(DWTCompose *cs, slice_buffer *sb,
401  IDWTELEM *temp,
402  int width, int height,
403  int stride_line)
404 {
405  int y = cs->y;
406 
407  IDWTELEM *b0 = cs->b0;
408  IDWTELEM *b1 = cs->b1;
409  IDWTELEM *b2 = slice_buffer_get_line(sb,
410  avpriv_mirror(y + 1, height - 1) *
411  stride_line);
412  IDWTELEM *b3 = slice_buffer_get_line(sb,
413  avpriv_mirror(y + 2, height - 1) *
414  stride_line);
415 
416  if (y + 1 < (unsigned)height && y < (unsigned)height) {
417  int x;
418 
419  for (x = 0; x < width; x++) {
420  b2[x] -= (b1[x] + b3[x] + 2) >> 2;
421  b1[x] += (b0[x] + b2[x]) >> 1;
422  }
423  } else {
424  if (y + 1 < (unsigned)height)
425  vertical_compose53iL0(b1, b2, b3, width);
426  if (y + 0 < (unsigned)height)
427  vertical_compose53iH0(b0, b1, b2, width);
428  }
429 
430  if (y - 1 < (unsigned)height)
431  horizontal_compose53i(b0, temp, width);
432  if (y + 0 < (unsigned)height)
433  horizontal_compose53i(b1, temp, width);
434 
435  cs->b0 = b2;
436  cs->b1 = b3;
437  cs->y += 2;
438 }
439 
440 static void spatial_compose53i_dy(DWTCompose *cs, IDWTELEM *buffer,
441  IDWTELEM *temp, int width, int height,
442  int stride)
443 {
444  int y = cs->y;
445  IDWTELEM *b0 = cs->b0;
446  IDWTELEM *b1 = cs->b1;
447  IDWTELEM *b2 = buffer + avpriv_mirror(y + 1, height - 1) * stride;
448  IDWTELEM *b3 = buffer + avpriv_mirror(y + 2, height - 1) * stride;
449 
450  if (y + 1 < (unsigned)height)
451  vertical_compose53iL0(b1, b2, b3, width);
452  if (y + 0 < (unsigned)height)
453  vertical_compose53iH0(b0, b1, b2, width);
454 
455  if (y - 1 < (unsigned)height)
456  horizontal_compose53i(b0, temp, width);
457  if (y + 0 < (unsigned)height)
458  horizontal_compose53i(b1, temp, width);
459 
460  cs->b0 = b2;
461  cs->b1 = b3;
462  cs->y += 2;
463 }
464 
465 void ff_snow_horizontal_compose97i(IDWTELEM *b, IDWTELEM *temp, int width)
466 {
467  const int w2 = (width + 1) >> 1;
468  int x;
469 
470  temp[0] = b[0] - ((3 * b[w2] + 2) >> 2);
471  for (x = 1; x < (width >> 1); x++) {
472  temp[2 * x] = b[x] - ((3 * (b[x + w2 - 1] + b[x + w2]) + 4) >> 3);
473  temp[2 * x - 1] = b[x + w2 - 1] - temp[2 * x - 2] - temp[2 * x];
474  }
475  if (width & 1) {
476  temp[2 * x] = b[x] - ((3 * b[x + w2 - 1] + 2) >> 2);
477  temp[2 * x - 1] = b[x + w2 - 1] - temp[2 * x - 2] - temp[2 * x];
478  } else
479  temp[2 * x - 1] = b[x + w2 - 1] - 2 * temp[2 * x - 2];
480 
481  b[0] = temp[0] + ((2 * temp[0] + temp[1] + 4) >> 3);
482  for (x = 2; x < width - 1; x += 2) {
483  b[x] = temp[x] + ((4 * temp[x] + temp[x - 1] + temp[x + 1] + 8) >> 4);
484  b[x - 1] = temp[x - 1] + ((3 * (b[x - 2] + b[x])) >> 1);
485  }
486  if (width & 1) {
487  b[x] = temp[x] + ((2 * temp[x] + temp[x - 1] + 4) >> 3);
488  b[x - 1] = temp[x - 1] + ((3 * (b[x - 2] + b[x])) >> 1);
489  } else
490  b[x - 1] = temp[x - 1] + 3 * b[x - 2];
491 }
492 
493 static void vertical_compose97iH0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
494  int width)
495 {
496  int i;
497 
498  for (i = 0; i < width; i++)
499  b1[i] += (W_AM * (b0[i] + b2[i]) + W_AO) >> W_AS;
500 }
501 
502 static void vertical_compose97iH1(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
503  int width)
504 {
505  int i;
506 
507  for (i = 0; i < width; i++)
508  b1[i] -= (W_CM * (b0[i] + b2[i]) + W_CO) >> W_CS;
509 }
510 
511 static void vertical_compose97iL0(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
512  int width)
513 {
514  int i;
515 
516  for (i = 0; i < width; i++)
517  b1[i] += (W_BM * (b0[i] + b2[i]) + 4 * b1[i] + W_BO) >> W_BS;
518 }
519 
520 static void vertical_compose97iL1(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
521  int width)
522 {
523  int i;
524 
525  for (i = 0; i < width; i++)
526  b1[i] -= (W_DM * (b0[i] + b2[i]) + W_DO) >> W_DS;
527 }
528 
529 void ff_snow_vertical_compose97i(IDWTELEM *b0, IDWTELEM *b1, IDWTELEM *b2,
530  IDWTELEM *b3, IDWTELEM *b4, IDWTELEM *b5,
531  int width)
532 {
533  int i;
534 
535  for (i = 0; i < width; i++) {
536  b4[i] -= (W_DM * (b3[i] + b5[i]) + W_DO) >> W_DS;
537  b3[i] -= (W_CM * (b2[i] + b4[i]) + W_CO) >> W_CS;
538  b2[i] += (W_BM * (b1[i] + b3[i]) + 4 * b2[i] + W_BO) >> W_BS;
539  b1[i] += (W_AM * (b0[i] + b2[i]) + W_AO) >> W_AS;
540  }
541 }
542 
543 static void spatial_compose97i_buffered_init(DWTCompose *cs, slice_buffer *sb,
544  int height, int stride_line)
545 {
546  cs->b0 = slice_buffer_get_line(sb, avpriv_mirror(-3 - 1, height - 1) * stride_line);
547  cs->b1 = slice_buffer_get_line(sb, avpriv_mirror(-3, height - 1) * stride_line);
548  cs->b2 = slice_buffer_get_line(sb, avpriv_mirror(-3 + 1, height - 1) * stride_line);
549  cs->b3 = slice_buffer_get_line(sb, avpriv_mirror(-3 + 2, height - 1) * stride_line);
550  cs->y = -3;
551 }
552 
553 static void spatial_compose97i_init(DWTCompose *cs, IDWTELEM *buffer, int height,
554  int stride)
555 {
556  cs->b0 = buffer + avpriv_mirror(-3 - 1, height - 1) * stride;
557  cs->b1 = buffer + avpriv_mirror(-3, height - 1) * stride;
558  cs->b2 = buffer + avpriv_mirror(-3 + 1, height - 1) * stride;
559  cs->b3 = buffer + avpriv_mirror(-3 + 2, height - 1) * stride;
560  cs->y = -3;
561 }
562 
563 static void spatial_compose97i_dy_buffered(SnowDWTContext *dsp, DWTCompose *cs,
564  slice_buffer * sb, IDWTELEM *temp,
565  int width, int height,
566  int stride_line)
567 {
568  int y = cs->y;
569 
570  IDWTELEM *b0 = cs->b0;
571  IDWTELEM *b1 = cs->b1;
572  IDWTELEM *b2 = cs->b2;
573  IDWTELEM *b3 = cs->b3;
574  IDWTELEM *b4 = slice_buffer_get_line(sb,
575  avpriv_mirror(y + 3, height - 1) *
576  stride_line);
577  IDWTELEM *b5 = slice_buffer_get_line(sb,
578  avpriv_mirror(y + 4, height - 1) *
579  stride_line);
580 
581  if (y > 0 && y + 4 < height) {
582  dsp->vertical_compose97i(b0, b1, b2, b3, b4, b5, width);
583  } else {
584  if (y + 3 < (unsigned)height)
585  vertical_compose97iL1(b3, b4, b5, width);
586  if (y + 2 < (unsigned)height)
587  vertical_compose97iH1(b2, b3, b4, width);
588  if (y + 1 < (unsigned)height)
589  vertical_compose97iL0(b1, b2, b3, width);
590  if (y + 0 < (unsigned)height)
591  vertical_compose97iH0(b0, b1, b2, width);
592  }
593 
594  if (y - 1 < (unsigned)height)
595  dsp->horizontal_compose97i(b0, temp, width);
596  if (y + 0 < (unsigned)height)
597  dsp->horizontal_compose97i(b1, temp, width);
598 
599  cs->b0 = b2;
600  cs->b1 = b3;
601  cs->b2 = b4;
602  cs->b3 = b5;
603  cs->y += 2;
604 }
605 
606 static void spatial_compose97i_dy(DWTCompose *cs, IDWTELEM *buffer,
607  IDWTELEM *temp, int width, int height,
608  int stride)
609 {
610  int y = cs->y;
611  IDWTELEM *b0 = cs->b0;
612  IDWTELEM *b1 = cs->b1;
613  IDWTELEM *b2 = cs->b2;
614  IDWTELEM *b3 = cs->b3;
615  IDWTELEM *b4 = buffer + avpriv_mirror(y + 3, height - 1) * stride;
616  IDWTELEM *b5 = buffer + avpriv_mirror(y + 4, height - 1) * stride;
617 
618  if (y + 3 < (unsigned)height)
619  vertical_compose97iL1(b3, b4, b5, width);
620  if (y + 2 < (unsigned)height)
621  vertical_compose97iH1(b2, b3, b4, width);
622  if (y + 1 < (unsigned)height)
623  vertical_compose97iL0(b1, b2, b3, width);
624  if (y + 0 < (unsigned)height)
625  vertical_compose97iH0(b0, b1, b2, width);
626 
627  if (y - 1 < (unsigned)height)
628  ff_snow_horizontal_compose97i(b0, temp, width);
629  if (y + 0 < (unsigned)height)
630  ff_snow_horizontal_compose97i(b1, temp, width);
631 
632  cs->b0 = b2;
633  cs->b1 = b3;
634  cs->b2 = b4;
635  cs->b3 = b5;
636  cs->y += 2;
637 }
638 
639 void ff_spatial_idwt_buffered_init(DWTCompose *cs, slice_buffer *sb, int width,
640  int height, int stride_line, int type,
641  int decomposition_count)
642 {
643  int level;
644  for (level = decomposition_count - 1; level >= 0; level--) {
645  switch (type) {
646  case DWT_97:
647  spatial_compose97i_buffered_init(cs + level, sb, height >> level,
648  stride_line << level);
649  break;
650  case DWT_53:
651  spatial_compose53i_buffered_init(cs + level, sb, height >> level,
652  stride_line << level);
653  break;
654  }
655  }
656 }
657 
658 void ff_spatial_idwt_buffered_slice(SnowDWTContext *dsp, DWTCompose *cs,
659  slice_buffer *slice_buf, IDWTELEM *temp,
660  int width, int height, int stride_line,
661  int type, int decomposition_count, int y)
662 {
663  const int support = type == 1 ? 3 : 5;
664  int level;
665  if (type == 2)
666  return;
667 
668  for (level = decomposition_count - 1; level >= 0; level--)
669  while (cs[level].y <= FFMIN((y >> level) + support, height >> level)) {
670  switch (type) {
671  case DWT_97:
672  spatial_compose97i_dy_buffered(dsp, cs + level, slice_buf, temp,
673  width >> level,
674  height >> level,
675  stride_line << level);
676  break;
677  case DWT_53:
678  spatial_compose53i_dy_buffered(cs + level, slice_buf, temp,
679  width >> level,
680  height >> level,
681  stride_line << level);
682  break;
683  }
684  }
685 }
686 
687 static void ff_spatial_idwt_init(DWTCompose *cs, IDWTELEM *buffer, int width,
688  int height, int stride, int type,
689  int decomposition_count)
690 {
691  int level;
692  for (level = decomposition_count - 1; level >= 0; level--) {
693  switch (type) {
694  case DWT_97:
695  spatial_compose97i_init(cs + level, buffer, height >> level,
696  stride << level);
697  break;
698  case DWT_53:
699  spatial_compose53i_init(cs + level, buffer, height >> level,
700  stride << level);
701  break;
702  }
703  }
704 }
705 
706 static void ff_spatial_idwt_slice(DWTCompose *cs, IDWTELEM *buffer,
707  IDWTELEM *temp, int width, int height,
708  int stride, int type,
709  int decomposition_count, int y)
710 {
711  const int support = type == 1 ? 3 : 5;
712  int level;
713  if (type == 2)
714  return;
715 
716  for (level = decomposition_count - 1; level >= 0; level--)
717  while (cs[level].y <= FFMIN((y >> level) + support, height >> level)) {
718  switch (type) {
719  case DWT_97:
720  spatial_compose97i_dy(cs + level, buffer, temp, width >> level,
721  height >> level, stride << level);
722  break;
723  case DWT_53:
724  spatial_compose53i_dy(cs + level, buffer, temp, width >> level,
725  height >> level, stride << level);
726  break;
727  }
728  }
729 }
730 
731 void ff_spatial_idwt(IDWTELEM *buffer, IDWTELEM *temp, int width, int height,
732  int stride, int type, int decomposition_count)
733 {
734  DWTCompose cs[MAX_DECOMPOSITIONS];
735  int y;
736  ff_spatial_idwt_init(cs, buffer, width, height, stride, type,
737  decomposition_count);
738  for (y = 0; y < height; y += 4)
739  ff_spatial_idwt_slice(cs, buffer, temp, width, height, stride, type,
740  decomposition_count, y);
741 }
742 
743 static inline int w_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size,
744  int w, int h, int type)
745 {
746  int s, i, j;
747  const int dec_count = w == 8 ? 3 : 4;
748  int tmp[32 * 32], tmp2[32];
749  int level, ori;
750  static const int scale[2][2][4][4] = {
751  {
752  { // 9/7 8x8 dec=3
753  { 268, 239, 239, 213 },
754  { 0, 224, 224, 152 },
755  { 0, 135, 135, 110 },
756  },
757  { // 9/7 16x16 or 32x32 dec=4
758  { 344, 310, 310, 280 },
759  { 0, 320, 320, 228 },
760  { 0, 175, 175, 136 },
761  { 0, 129, 129, 102 },
762  }
763  },
764  {
765  { // 5/3 8x8 dec=3
766  { 275, 245, 245, 218 },
767  { 0, 230, 230, 156 },
768  { 0, 138, 138, 113 },
769  },
770  { // 5/3 16x16 or 32x32 dec=4
771  { 352, 317, 317, 286 },
772  { 0, 328, 328, 233 },
773  { 0, 180, 180, 140 },
774  { 0, 132, 132, 105 },
775  }
776  }
777  };
778 
779  for (i = 0; i < h; i++) {
780  for (j = 0; j < w; j += 4) {
781  tmp[32 * i + j + 0] = (pix1[j + 0] - pix2[j + 0]) << 4;
782  tmp[32 * i + j + 1] = (pix1[j + 1] - pix2[j + 1]) << 4;
783  tmp[32 * i + j + 2] = (pix1[j + 2] - pix2[j + 2]) << 4;
784  tmp[32 * i + j + 3] = (pix1[j + 3] - pix2[j + 3]) << 4;
785  }
786  pix1 += line_size;
787  pix2 += line_size;
788  }
789 
790  ff_spatial_dwt(tmp, tmp2, w, h, 32, type, dec_count);
791 
792  s = 0;
793  av_assert1(w == h);
794  for (level = 0; level < dec_count; level++)
795  for (ori = level ? 1 : 0; ori < 4; ori++) {
796  int size = w >> (dec_count - level);
797  int sx = (ori & 1) ? size : 0;
798  int stride = 32 << (dec_count - level);
799  int sy = (ori & 2) ? stride >> 1 : 0;
800 
801  for (i = 0; i < size; i++)
802  for (j = 0; j < size; j++) {
803  int v = tmp[sx + sy + i * stride + j] *
804  scale[type][dec_count - 3][level][ori];
805  s += FFABS(v);
806  }
807  }
808  av_assert1(s >= 0);
809  return s >> 9;
810 }
811 
812 static int w53_8_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
813 {
814  return w_c(v, pix1, pix2, line_size, 8, h, 1);
815 }
816 
817 static int w97_8_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
818 {
819  return w_c(v, pix1, pix2, line_size, 8, h, 0);
820 }
821 
822 static int w53_16_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
823 {
824  return w_c(v, pix1, pix2, line_size, 16, h, 1);
825 }
826 
827 static int w97_16_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
828 {
829  return w_c(v, pix1, pix2, line_size, 16, h, 0);
830 }
831 
832 int ff_w53_32_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
833 {
834  return w_c(v, pix1, pix2, line_size, 32, h, 1);
835 }
836 
837 int ff_w97_32_c(struct MpegEncContext *v, uint8_t *pix1, uint8_t *pix2, ptrdiff_t line_size, int h)
838 {
839  return w_c(v, pix1, pix2, line_size, 32, h, 0);
840 }
841 
842 void ff_dsputil_init_dwt(MECmpContext *c)
843 {
844  c->w53[0] = w53_16_c;
845  c->w53[1] = w53_8_c;
846  c->w97[0] = w97_16_c;
847  c->w97[1] = w97_8_c;
848 }
849 
850 void ff_dwt_init(SnowDWTContext *c)
851 {
852  c->vertical_compose97i = ff_snow_vertical_compose97i;
853  c->horizontal_compose97i = ff_snow_horizontal_compose97i;
854  c->inner_add_yblock = ff_snow_inner_add_yblock;
855 
856  if (HAVE_MMX)
857  ff_dwt_init_x86(c);
858 }
859 
860 
Generated on Sun Jan 25 2015 19:21:16 for FFmpeg by   doxygen 1.8.2