filter_c.c 15.4 KB
Newer Older
John Koleszar's avatar
John Koleszar committed
1
/*
2
 *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
John Koleszar's avatar
John Koleszar committed
3
 *
4
 *  Use of this source code is governed by a BSD-style license
5
6
 *  that can be found in the LICENSE file in the root of the source
 *  tree. An additional intellectual property rights grant can be found
7
 *  in the file PATENTS.  All contributing project authors may
8
 *  be found in the AUTHORS file in the root of the source tree.
John Koleszar's avatar
John Koleszar committed
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
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
 */


#include <stdlib.h>

#define BLOCK_HEIGHT_WIDTH 4
#define VP8_FILTER_WEIGHT 128
#define VP8_FILTER_SHIFT  7


static const int bilinear_filters[8][2] =
{
    { 128,   0 },
    { 112,  16 },
    {  96,  32 },
    {  80,  48 },
    {  64,  64 },
    {  48,  80 },
    {  32,  96 },
    {  16, 112 }
};


static const short sub_pel_filters[8][6] =
{

    { 0,  0,  128,    0,   0,  0 },         // note that 1/8 pel positions are just as per alpha -0.5 bicubic
    { 0, -6,  123,   12,  -1,  0 },
    { 2, -11, 108,   36,  -8,  1 },         // New 1/4 pel 6 tap filter
    { 0, -9,   93,   50,  -6,  0 },
    { 3, -16,  77,   77, -16,  3 },         // New 1/2 pel 6 tap filter
    { 0, -6,   50,   93,  -9,  0 },
    { 1, -8,   36,  108, -11,  2 },         // New 1/4 pel 6 tap filter
    { 0, -1,   12,  123,  -6,  0 },



};

void vp8_filter_block2d_first_pass
(
    unsigned char *src_ptr,
    int *output_ptr,
    unsigned int src_pixels_per_line,
    unsigned int pixel_step,
    unsigned int output_height,
    unsigned int output_width,
    const short *vp8_filter
)
{
    unsigned int i, j;
    int  Temp;

    for (i = 0; i < output_height; i++)
    {
        for (j = 0; j < output_width; j++)
        {
            Temp = ((int)src_ptr[-2 * (int)pixel_step] * vp8_filter[0]) +
                   ((int)src_ptr[-1 * (int)pixel_step] * vp8_filter[1]) +
                   ((int)src_ptr[0]                 * vp8_filter[2]) +
                   ((int)src_ptr[pixel_step]         * vp8_filter[3]) +
                   ((int)src_ptr[2*pixel_step]       * vp8_filter[4]) +
                   ((int)src_ptr[3*pixel_step]       * vp8_filter[5]) +
                   (VP8_FILTER_WEIGHT >> 1);      // Rounding

            // Normalize back to 0-255
            Temp = Temp >> VP8_FILTER_SHIFT;

            if (Temp < 0)
                Temp = 0;
            else if (Temp > 255)
                Temp = 255;

            output_ptr[j] = Temp;
            src_ptr++;
        }

        // Next row...
        src_ptr    += src_pixels_per_line - output_width;
        output_ptr += output_width;
    }
}

void vp8_filter_block2d_second_pass
(
    int *src_ptr,
    unsigned char *output_ptr,
    int output_pitch,
    unsigned int src_pixels_per_line,
    unsigned int pixel_step,
    unsigned int output_height,
    unsigned int output_width,
    const short *vp8_filter
)
{
    unsigned int i, j;
    int  Temp;

    for (i = 0; i < output_height; i++)
    {
        for (j = 0; j < output_width; j++)
        {
            // Apply filter
            Temp = ((int)src_ptr[-2 * (int)pixel_step] * vp8_filter[0]) +
                   ((int)src_ptr[-1 * (int)pixel_step] * vp8_filter[1]) +
                   ((int)src_ptr[0]                 * vp8_filter[2]) +
                   ((int)src_ptr[pixel_step]         * vp8_filter[3]) +
                   ((int)src_ptr[2*pixel_step]       * vp8_filter[4]) +
                   ((int)src_ptr[3*pixel_step]       * vp8_filter[5]) +
                   (VP8_FILTER_WEIGHT >> 1);   // Rounding

            // Normalize back to 0-255
            Temp = Temp >> VP8_FILTER_SHIFT;

            if (Temp < 0)
                Temp = 0;
            else if (Temp > 255)
                Temp = 255;

            output_ptr[j] = (unsigned char)Temp;
            src_ptr++;
        }

        // Start next row
        src_ptr    += src_pixels_per_line - output_width;
        output_ptr += output_pitch;
    }
}


void vp8_filter_block2d
(
    unsigned char  *src_ptr,
    unsigned char  *output_ptr,
    unsigned int src_pixels_per_line,
    int output_pitch,
    const short  *HFilter,
    const short  *VFilter
)
{
    int FData[9*4]; // Temp data bufffer used in filtering

    // First filter 1-D horizontally...
    vp8_filter_block2d_first_pass(src_ptr - (2 * src_pixels_per_line), FData, src_pixels_per_line, 1, 9, 4, HFilter);

    // then filter verticaly...
    vp8_filter_block2d_second_pass(FData + 8, output_ptr, output_pitch, 4, 4, 4, 4, VFilter);
}


void vp8_block_variation_c
(
    unsigned char  *src_ptr,
    int   src_pixels_per_line,
    int *HVar,
    int *VVar
)
{
    int i, j;
    unsigned char *Ptr = src_ptr;

    for (i = 0; i < 4; i++)
    {
        for (j = 0; j < 4; j++)
        {
            *HVar += abs((int)Ptr[j] - (int)Ptr[j+1]);
            *VVar += abs((int)Ptr[j] - (int)Ptr[j+src_pixels_per_line]);
        }

        Ptr += src_pixels_per_line;
    }
}




void vp8_sixtap_predict_c
(
    unsigned char  *src_ptr,
    int   src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int dst_pitch
)
{
    const short  *HFilter;
    const short  *VFilter;

    HFilter = sub_pel_filters[xoffset];   // 6 tap
    VFilter = sub_pel_filters[yoffset];   // 6 tap

    vp8_filter_block2d(src_ptr, dst_ptr, src_pixels_per_line, dst_pitch, HFilter, VFilter);
}
void vp8_sixtap_predict8x8_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const short  *HFilter;
    const short  *VFilter;
    int FData[13*16];   // Temp data bufffer used in filtering

    HFilter = sub_pel_filters[xoffset];   // 6 tap
    VFilter = sub_pel_filters[yoffset];   // 6 tap

    // First filter 1-D horizontally...
    vp8_filter_block2d_first_pass(src_ptr - (2 * src_pixels_per_line), FData, src_pixels_per_line, 1, 13, 8, HFilter);


    // then filter verticaly...
    vp8_filter_block2d_second_pass(FData + 16, dst_ptr, dst_pitch, 8, 8, 8, 8, VFilter);

}

void vp8_sixtap_predict8x4_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const short  *HFilter;
    const short  *VFilter;
    int FData[13*16];   // Temp data bufffer used in filtering

    HFilter = sub_pel_filters[xoffset];   // 6 tap
    VFilter = sub_pel_filters[yoffset];   // 6 tap

    // First filter 1-D horizontally...
    vp8_filter_block2d_first_pass(src_ptr - (2 * src_pixels_per_line), FData, src_pixels_per_line, 1, 9, 8, HFilter);


    // then filter verticaly...
    vp8_filter_block2d_second_pass(FData + 16, dst_ptr, dst_pitch, 8, 8, 4, 8, VFilter);

}

void vp8_sixtap_predict16x16_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const short  *HFilter;
    const short  *VFilter;
    int FData[21*24];   // Temp data bufffer used in filtering


    HFilter = sub_pel_filters[xoffset];   // 6 tap
    VFilter = sub_pel_filters[yoffset];   // 6 tap

    // First filter 1-D horizontally...
    vp8_filter_block2d_first_pass(src_ptr - (2 * src_pixels_per_line), FData, src_pixels_per_line, 1, 21, 16, HFilter);

    // then filter verticaly...
    vp8_filter_block2d_second_pass(FData + 32, dst_ptr, dst_pitch, 16, 16, 16, 16, VFilter);

}


/****************************************************************************
 *
 *  ROUTINE       : filter_block2d_bil_first_pass
 *
 *  INPUTS        : UINT8  *src_ptr          : Pointer to source block.
 *                  UINT32 src_pixels_per_line : Stride of input block.
 *                  UINT32 pixel_step        : Offset between filter input samples (see notes).
 *                  UINT32 output_height     : Input block height.
 *                  UINT32 output_width      : Input block width.
 *                  INT32  *vp8_filter          : Array of 2 bi-linear filter taps.
 *
 *  OUTPUTS       : INT32 *output_ptr        : Pointer to filtered block.
 *
 *  RETURNS       : void
 *
 *  FUNCTION      : Applies a 1-D 2-tap bi-linear filter to the source block in
 *                  either horizontal or vertical direction to produce the
 *                  filtered output block. Used to implement first-pass
 *                  of 2-D separable filter.
 *
 *  SPECIAL NOTES : Produces INT32 output to retain precision for next pass.
 *                  Two filter taps should sum to VP8_FILTER_WEIGHT.
 *                  pixel_step defines whether the filter is applied
 *                  horizontally (pixel_step=1) or vertically (pixel_step=stride).
 *                  It defines the offset required to move from one input
 *                  to the next.
 *
 ****************************************************************************/
void vp8_filter_block2d_bil_first_pass
(
    unsigned char *src_ptr,
    unsigned short *output_ptr,
    unsigned int src_pixels_per_line,
    int pixel_step,
    unsigned int output_height,
    unsigned int output_width,
    const int *vp8_filter
)
{
    unsigned int i, j;

    for (i = 0; i < output_height; i++)
    {
        for (j = 0; j < output_width; j++)
        {
            // Apply bilinear filter
            output_ptr[j] = (((int)src_ptr[0]          * vp8_filter[0]) +
                             ((int)src_ptr[pixel_step] * vp8_filter[1]) +
                             (VP8_FILTER_WEIGHT / 2)) >> VP8_FILTER_SHIFT;
            src_ptr++;
        }

        // Next row...
        src_ptr    += src_pixels_per_line - output_width;
        output_ptr += output_width;
    }
}

/****************************************************************************
 *
 *  ROUTINE       : filter_block2d_bil_second_pass
 *
 *  INPUTS        : INT32  *src_ptr          : Pointer to source block.
 *                  UINT32 src_pixels_per_line : Stride of input block.
 *                  UINT32 pixel_step        : Offset between filter input samples (see notes).
 *                  UINT32 output_height     : Input block height.
 *                  UINT32 output_width      : Input block width.
 *                  INT32  *vp8_filter          : Array of 2 bi-linear filter taps.
 *
 *  OUTPUTS       : UINT16 *output_ptr       : Pointer to filtered block.
 *
 *  RETURNS       : void
 *
 *  FUNCTION      : Applies a 1-D 2-tap bi-linear filter to the source block in
 *                  either horizontal or vertical direction to produce the
 *                  filtered output block. Used to implement second-pass
 *                  of 2-D separable filter.
 *
 *  SPECIAL NOTES : Requires 32-bit input as produced by filter_block2d_bil_first_pass.
 *                  Two filter taps should sum to VP8_FILTER_WEIGHT.
 *                  pixel_step defines whether the filter is applied
 *                  horizontally (pixel_step=1) or vertically (pixel_step=stride).
 *                  It defines the offset required to move from one input
 *                  to the next.
 *
 ****************************************************************************/
void vp8_filter_block2d_bil_second_pass
(
    unsigned short *src_ptr,
    unsigned char  *output_ptr,
    int output_pitch,
    unsigned int  src_pixels_per_line,
    unsigned int  pixel_step,
    unsigned int  output_height,
    unsigned int  output_width,
    const int *vp8_filter
)
{
    unsigned int  i, j;
    int  Temp;

    for (i = 0; i < output_height; i++)
    {
        for (j = 0; j < output_width; j++)
        {
            // Apply filter
            Temp = ((int)src_ptr[0]         * vp8_filter[0]) +
                   ((int)src_ptr[pixel_step] * vp8_filter[1]) +
                   (VP8_FILTER_WEIGHT / 2);
            output_ptr[j] = (unsigned int)(Temp >> VP8_FILTER_SHIFT);
            src_ptr++;
        }

        // Next row...
        src_ptr    += src_pixels_per_line - output_width;
        output_ptr += output_pitch;
    }
}


/****************************************************************************
 *
 *  ROUTINE       : filter_block2d_bil
 *
 *  INPUTS        : UINT8  *src_ptr          : Pointer to source block.
 *                  UINT32 src_pixels_per_line : Stride of input block.
 *                  INT32  *HFilter         : Array of 2 horizontal filter taps.
 *                  INT32  *VFilter         : Array of 2 vertical filter taps.
 *
 *  OUTPUTS       : UINT16 *output_ptr       : Pointer to filtered block.
 *
 *  RETURNS       : void
 *
 *  FUNCTION      : 2-D filters an input block by applying a 2-tap
 *                  bi-linear filter horizontally followed by a 2-tap
 *                  bi-linear filter vertically on the result.
 *
 *  SPECIAL NOTES : The largest block size can be handled here is 16x16
 *
 ****************************************************************************/
void vp8_filter_block2d_bil
(
    unsigned char *src_ptr,
    unsigned char *output_ptr,
    unsigned int   src_pixels_per_line,
    unsigned int   dst_pitch,
    const int      *HFilter,
    const int      *VFilter,
    int            Width,
    int            Height
)
{

    unsigned short FData[17*16];    // Temp data bufffer used in filtering

    // First filter 1-D horizontally...
    vp8_filter_block2d_bil_first_pass(src_ptr, FData, src_pixels_per_line, 1, Height + 1, Width, HFilter);

    // then 1-D vertically...
    vp8_filter_block2d_bil_second_pass(FData, output_ptr, dst_pitch, Width, Width, Height, Width, VFilter);
}


void vp8_bilinear_predict4x4_c
(
    unsigned char  *src_ptr,
    int   src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int dst_pitch
)
{
    const int  *HFilter;
    const int  *VFilter;

    HFilter = bilinear_filters[xoffset];
    VFilter = bilinear_filters[yoffset];
#if 0
    {
        int i;
        unsigned char temp1[16];
        unsigned char temp2[16];

        bilinear_predict4x4_mmx(src_ptr, src_pixels_per_line, xoffset, yoffset, temp1, 4);
        vp8_filter_block2d_bil(src_ptr, temp2, src_pixels_per_line, 4, HFilter, VFilter, 4, 4);

        for (i = 0; i < 16; i++)
        {
            if (temp1[i] != temp2[i])
            {
                bilinear_predict4x4_mmx(src_ptr, src_pixels_per_line, xoffset, yoffset, temp1, 4);
                vp8_filter_block2d_bil(src_ptr, temp2, src_pixels_per_line, 4, HFilter, VFilter, 4, 4);
            }
        }
    }
#endif
    vp8_filter_block2d_bil(src_ptr, dst_ptr, src_pixels_per_line, dst_pitch, HFilter, VFilter, 4, 4);

}

void vp8_bilinear_predict8x8_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const int  *HFilter;
    const int  *VFilter;

    HFilter = bilinear_filters[xoffset];
    VFilter = bilinear_filters[yoffset];

    vp8_filter_block2d_bil(src_ptr, dst_ptr, src_pixels_per_line, dst_pitch, HFilter, VFilter, 8, 8);

}

void vp8_bilinear_predict8x4_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const int  *HFilter;
    const int  *VFilter;

    HFilter = bilinear_filters[xoffset];
    VFilter = bilinear_filters[yoffset];

    vp8_filter_block2d_bil(src_ptr, dst_ptr, src_pixels_per_line, dst_pitch, HFilter, VFilter, 8, 4);

}

void vp8_bilinear_predict16x16_c
(
    unsigned char  *src_ptr,
    int  src_pixels_per_line,
    int  xoffset,
    int  yoffset,
    unsigned char *dst_ptr,
    int  dst_pitch
)
{
    const int  *HFilter;
    const int  *VFilter;

    HFilter = bilinear_filters[xoffset];
    VFilter = bilinear_filters[yoffset];

    vp8_filter_block2d_bil(src_ptr, dst_ptr, src_pixels_per_line, dst_pitch, HFilter, VFilter, 16, 16);
}