Parent Directory | Revision Log
Revision 1.2 - (view) (download)
1 : | Isibaar | 1.1 | /***************************************************************************** |
2 : | * | ||
3 : | * XVID MPEG-4 VIDEO CODEC | ||
4 : | * - PSNR-HVS-M plugin: computes the PSNR-HVS-M metric - | ||
5 : | * | ||
6 : | * Copyright(C) 2010 Michael Militzer <michael@xvid.org> | ||
7 : | * | ||
8 : | * This program is free software ; you can redistribute it and/or modify | ||
9 : | * it under the terms of the GNU General Public License as published by | ||
10 : | * the Free Software Foundation ; either version 2 of the License, or | ||
11 : | * (at your option) any later version. | ||
12 : | * | ||
13 : | * This program is distributed in the hope that it will be useful, | ||
14 : | * but WITHOUT ANY WARRANTY ; without even the implied warranty of | ||
15 : | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | ||
16 : | * GNU General Public License for more details. | ||
17 : | * | ||
18 : | * You should have received a copy of the GNU General Public License | ||
19 : | * along with this program ; if not, write to the Free Software | ||
20 : | * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA | ||
21 : | * | ||
22 : | Isibaar | 1.2 | * $Id: plugin_psnrhvsm.c,v 1.1 2010/10/10 19:19:46 Isibaar Exp $ |
23 : | Isibaar | 1.1 | * |
24 : | ****************************************************************************/ | ||
25 : | |||
26 : | /***************************************************************************** | ||
27 : | * | ||
28 : | * The PSNR-HVS-M metric is described in the following paper: | ||
29 : | * | ||
30 : | * "On between-coefficient contrast masking of DCT basis functions", by | ||
31 : | * N. Ponomarenko, F. Silvestri, K. Egiazarian, M. Carli, J. Astola, V. Lukin, | ||
32 : | * in Proceedings of the Third International Workshop on Video Processing and | ||
33 : | * Quality Metrics for Consumer Electronics VPQM-07, January, 2007, 4 p. | ||
34 : | * | ||
35 : | * http://www.ponomarenko.info/psnrhvsm.htm | ||
36 : | * | ||
37 : | ****************************************************************************/ | ||
38 : | |||
39 : | #include <stdlib.h> | ||
40 : | #include <stdio.h> | ||
41 : | #include <math.h> | ||
42 : | #include "../portab.h" | ||
43 : | #include "../xvid.h" | ||
44 : | #include "../dct/fdct.h" | ||
45 : | #include "../image/image.h" | ||
46 : | #include "../utils/mem_transfer.h" | ||
47 : | #include "../utils/emms.h" | ||
48 : | |||
49 : | typedef struct { | ||
50 : | |||
51 : | uint64_t mse_sum; /* for avrg psnrhvsm */ | ||
52 : | long frame_cnt; | ||
53 : | |||
54 : | } psnrhvsm_data_t; /* internal plugin data */ | ||
55 : | |||
56 : | static const float CSF_Coeff[64] = | ||
57 : | { 1.608443f, 2.339554f, 2.573509f, 1.608443f, 1.072295f, 0.643377f, 0.504610f, 0.421887f, | ||
58 : | 2.144591f, 2.144591f, 1.838221f, 1.354478f, 0.989811f, 0.443708f, 0.428918f, 0.467911f, | ||
59 : | 1.838221f, 1.979622f, 1.608443f, 1.072295f, 0.643377f, 0.451493f, 0.372972f, 0.459555f, | ||
60 : | 1.838221f, 1.513829f, 1.169777f, 0.887417f, 0.504610f, 0.295806f, 0.321689f, 0.415082f, | ||
61 : | 1.429727f, 1.169777f, 0.695543f, 0.459555f, 0.378457f, 0.236102f, 0.249855f, 0.334222f, | ||
62 : | 1.072295f, 0.735288f, 0.467911f, 0.402111f, 0.317717f, 0.247453f, 0.227744f, 0.279729f, | ||
63 : | 0.525206f, 0.402111f, 0.329937f, 0.295806f, 0.249855f, 0.212687f, 0.214459f, 0.254803f, | ||
64 : | 0.357432f, 0.279729f, 0.270896f, 0.262603f, 0.229778f, 0.257351f, 0.249855f, 0.259950f | ||
65 : | }; | ||
66 : | |||
67 : | static const float Mask_Coeff[64] = | ||
68 : | { 0.000000f, 0.826446f, 1.000000f, 0.390625f, 0.173611f, 0.062500f, 0.038447f, 0.026874f, | ||
69 : | 0.694444f, 0.694444f, 0.510204f, 0.277008f, 0.147929f, 0.029727f, 0.027778f, 0.033058f, | ||
70 : | 0.510204f, 0.591716f, 0.390625f, 0.173611f, 0.062500f, 0.030779f, 0.021004f, 0.031888f, | ||
71 : | 0.510204f, 0.346021f, 0.206612f, 0.118906f, 0.038447f, 0.013212f, 0.015625f, 0.026015f, | ||
72 : | 0.308642f, 0.206612f, 0.073046f, 0.031888f, 0.021626f, 0.008417f, 0.009426f, 0.016866f, | ||
73 : | 0.173611f, 0.081633f, 0.033058f, 0.024414f, 0.015242f, 0.009246f, 0.007831f, 0.011815f, | ||
74 : | 0.041649f, 0.024414f, 0.016437f, 0.013212f, 0.009426f, 0.006830f, 0.006944f, 0.009803f, | ||
75 : | 0.019290f, 0.011815f, 0.011080f, 0.010412f, 0.007972f, 0.010000f, 0.009426f, 0.010203f | ||
76 : | }; | ||
77 : | |||
78 : | Isibaar | 1.2 | #if 0 /* Floating-point implementation */ |
79 : | Isibaar | 1.1 | |
80 : | static uint32_t Calc_MSE_H(int16_t *DCT_A, int16_t *DCT_B, uint8_t *IMG_A, uint8_t *IMG_B, int stride) | ||
81 : | { | ||
82 : | int x, y, i, j; | ||
83 : | uint32_t Global_A, Global_B, Sum_A = 0, Sum_B = 0; | ||
84 : | uint32_t Local[8] = {0, 0, 0, 0, 0, 0, 0, 0}; | ||
85 : | uint32_t Local_Square[8] = {0, 0, 0, 0, 0, 0, 0, 0}; | ||
86 : | float MASK_A = 0.f, MASK_B = 0.f; | ||
87 : | Isibaar | 1.2 | float Mult1 = 1.f, Mult2 = 1.f; |
88 : | Isibaar | 1.1 | uint32_t MSE_H = 0; |
89 : | |||
90 : | /* Step 1: Calculate CSF weighted energy of DCT coefficients */ | ||
91 : | for (y = 0; y < 8; y++) { | ||
92 : | for (x = 0; x < 8; x++) { | ||
93 : | MASK_A += (float)(DCT_A[y*8 + x]*DCT_A[y*8 + x])*Mask_Coeff[y*8 + x]; | ||
94 : | MASK_B += (float)(DCT_B[y*8 + x]*DCT_B[y*8 + x])*Mask_Coeff[y*8 + x]; | ||
95 : | } | ||
96 : | } | ||
97 : | |||
98 : | /* Step 2: Determine local variances compared to entire block variance */ | ||
99 : | for (y = 0; y < 2; y++) { | ||
100 : | for (x = 0; x < 2; x++) { | ||
101 : | Isibaar | 1.2 | for (j = 0; j < 4; j++) { |
102 : | for (i = 0; i < 4; i++) { | ||
103 : | uint8_t A = IMG_A[(y*4+j)*stride + 4*x + i]; | ||
104 : | uint8_t B = IMG_B[(y*4+j)*stride + 4*x + i]; | ||
105 : | |||
106 : | Local[y*2 + x] += A; | ||
107 : | Local[y*2 + x + 4] += B; | ||
108 : | Local_Square[y*2 + x] += A*A; | ||
109 : | Local_Square[y*2 + x + 4] += B*B; | ||
110 : | } | ||
111 : | Isibaar | 1.1 | } |
112 : | } | ||
113 : | } | ||
114 : | |||
115 : | Global_A = Local[0] + Local[1] + Local[2] + Local[3]; | ||
116 : | Global_B = Local[4] + Local[5] + Local[6] + Local[7]; | ||
117 : | |||
118 : | for (i = 0; i < 8; i++) | ||
119 : | Local[i] = (Local_Square[i]<<4) - (Local[i]*Local[i]); /* 16*Var(Di) */ | ||
120 : | |||
121 : | Local_Square[0] += (Local_Square[1] + Local_Square[2] + Local_Square[3]); | ||
122 : | Local_Square[4] += (Local_Square[5] + Local_Square[6] + Local_Square[7]); | ||
123 : | |||
124 : | Global_A = (Local_Square[0]<<6) - Global_A*Global_A; /* 64*Var(D) */ | ||
125 : | Global_B = (Local_Square[4]<<6) - Global_B*Global_B; /* 64*Var(D) */ | ||
126 : | |||
127 : | /* Step 3: Calculate contrast masking threshold */ | ||
128 : | Isibaar | 1.2 | if (Global_A) |
129 : | Mult1 = (float)(Local[0]+Local[1]+Local[2]+Local[3])/((float)(Global_A)/4.f); | ||
130 : | |||
131 : | if (Global_B) | ||
132 : | Mult2 = (float)(Local[4]+Local[5]+Local[6]+Local[7])/((float)(Global_B)/4.f); | ||
133 : | |||
134 : | MASK_A = (float)sqrt(MASK_A * Mult1) / 32.f; | ||
135 : | MASK_B = (float)sqrt(MASK_B * Mult2) / 32.f; | ||
136 : | Isibaar | 1.1 | |
137 : | if (MASK_B > MASK_A) MASK_A = MASK_B; /* MAX(MASK_A, MASK_B) */ | ||
138 : | |||
139 : | /* Step 4: Calculate MSE of DCT coeffs reduced by masking effect */ | ||
140 : | for (j = 0; j < 8; j++) { | ||
141 : | for (i = 0; i < 8; i++) { | ||
142 : | float u = (float)abs(DCT_A[j*8 + i] - DCT_B[j*8 + i]); | ||
143 : | |||
144 : | if ((i|j)>0) { | ||
145 : | if (u < (MASK_A / Mask_Coeff[j*8 + i])) | ||
146 : | u = 0; /* The error is not perceivable */ | ||
147 : | else | ||
148 : | u -= (MASK_A / Mask_Coeff[j*8 + i]); | ||
149 : | } | ||
150 : | |||
151 : | MSE_H += (uint32_t) ((256.f*(u * CSF_Coeff[j*8 + i])*(u * CSF_Coeff[j*8 + i])) + 0.5f); | ||
152 : | } | ||
153 : | } | ||
154 : | return MSE_H; /* Fixed-point value right-shifted by eight */ | ||
155 : | } | ||
156 : | |||
157 : | #else /* First draft of a fixed-point implementation. | ||
158 : | Might serve as a template for MMX/SSE code */ | ||
159 : | |||
160 : | static const uint16_t iMask_Coeff[64] = | ||
161 : | { 0, 59577, 65535, 40959, 27306, 16384, 12850, 10743, | ||
162 : | 54612, 54612, 46811, 34492, 25206, 11299, 10923, 11915, | ||
163 : | 46811, 50412, 40959, 27306, 16384, 11497, 9498, 11703, | ||
164 : | 46811, 38550, 29789, 22598, 12850, 7533, 8192, 10570, | ||
165 : | 36408, 29789, 17712, 11703, 9637, 6012, 6363, 8511, | ||
166 : | 27306, 18724, 11915, 10240, 8091, 6302, 5799, 7123, | ||
167 : | 13374, 10240, 8402, 7533, 6363, 5416, 5461, 6489, | ||
168 : | 9102, 7123, 6898, 6687, 5851, 6553, 6363, 6620 | ||
169 : | }; | ||
170 : | |||
171 : | static const uint16_t Inv_iMask_Coeff[64] = | ||
172 : | { 0, 310, 256, 655, 1475, 4096, 6659, 9526, | ||
173 : | 369, 369, 502, 924, 1731, 8612, 9216, 7744, | ||
174 : | 502, 433, 655, 1475, 4096, 8317, 12188, 8028, | ||
175 : | 502, 740, 1239, 2153, 6659, 19376, 16384, 9840, | ||
176 : | 829, 1239, 3505, 8028, 11838, 30415, 27159, 15178, | ||
177 : | 1475, 3136, 7744, 10486, 16796, 27688, 32691, 21667, | ||
178 : | 6147, 10486, 15575, 19376, 27159, 37482, 36866, 26114, | ||
179 : | 13271, 21667, 23105, 24587, 32112, 25600, 27159, 25091 | ||
180 : | }; | ||
181 : | |||
182 : | static const uint16_t iCSF_Coeff[64] = | ||
183 : | { 1647, 2396, 2635, 1647, 1098, 659, 517, 432, | ||
184 : | 2196, 2196, 1882, 1387, 1014, 454, 439, 479, | ||
185 : | 1882, 2027, 1647, 1098, 659, 462, 382, 471, | ||
186 : | 1882, 1550, 1198, 909, 517, 303, 329, 425, | ||
187 : | 1464, 1198, 712, 471, 388, 242, 256, 342, | ||
188 : | 1098, 753, 479, 412, 325, 253, 233, 286, | ||
189 : | 538, 412, 338, 303, 256, 218, 220, 261, | ||
190 : | 366, 286, 277, 269, 235, 264, 256, 266 | ||
191 : | }; | ||
192 : | |||
193 : | Isibaar | 1.2 | static __inline uint32_t isqrt(unsigned long n) |
194 : | { | ||
195 : | uint32_t c = 0x8000; | ||
196 : | uint32_t g = 0x8000; | ||
197 : | |||
198 : | for(;;) { | ||
199 : | if(g*g > n) | ||
200 : | g ^= c; | ||
201 : | c >>= 1; | ||
202 : | if(c == 0) | ||
203 : | return g; | ||
204 : | g |= c; | ||
205 : | } | ||
206 : | } | ||
207 : | |||
208 : | Isibaar | 1.1 | static uint32_t Calc_MSE_H(int16_t *DCT_A, int16_t *DCT_B, uint8_t *IMG_A, uint8_t *IMG_B, int stride) |
209 : | { | ||
210 : | int x, y, i, j; | ||
211 : | uint32_t Global_A, Global_B, Sum_A = 0, Sum_B = 0; | ||
212 : | uint32_t Local[8] = {0, 0, 0, 0, 0, 0, 0, 0}; | ||
213 : | uint32_t Local_Square[8] = {0, 0, 0, 0, 0, 0, 0, 0}; | ||
214 : | uint32_t MASK; | ||
215 : | uint32_t MSE_H = 0; | ||
216 : | |||
217 : | /* Step 1: Calculate CSF weighted energy of DCT coefficients */ | ||
218 : | for (y = 0; y < 8; y++) { | ||
219 : | for (x = 0; x < 8; x++) { | ||
220 : | uint16_t A = (abs(DCT_A[y*8 + x]) * iMask_Coeff[y*8 + x] + 4096) >> 13; | ||
221 : | uint16_t B = (abs(DCT_B[y*8 + x]) * iMask_Coeff[y*8 + x] + 4096) >> 13; | ||
222 : | |||
223 : | Sum_A += ((A*A) >> 3); /* PMADDWD */ | ||
224 : | Sum_B += ((B*B) >> 3); | ||
225 : | } | ||
226 : | } | ||
227 : | |||
228 : | /* Step 2: Determine local variances compared to entire block variance */ | ||
229 : | for (y = 0; y < 2; y++) { | ||
230 : | for (x = 0; x < 2; x++) { | ||
231 : | Isibaar | 1.2 | for (j = 0; j < 4; j++) { |
232 : | for (i = 0; i < 4; i++) { | ||
233 : | uint8_t A = IMG_A[(y*4+j)*stride + 4*x + i]; | ||
234 : | uint8_t B = IMG_B[(y*4+j)*stride + 4*x + i]; | ||
235 : | |||
236 : | Local[y*2 + x] += A; | ||
237 : | Local[y*2 + x + 4] += B; | ||
238 : | Local_Square[y*2 + x] += A*A; | ||
239 : | Local_Square[y*2 + x + 4] += B*B; | ||
240 : | } | ||
241 : | Isibaar | 1.1 | } |
242 : | } | ||
243 : | } | ||
244 : | |||
245 : | Global_A = Local[0] + Local[1] + Local[2] + Local[3]; | ||
246 : | Global_B = Local[4] + Local[5] + Local[6] + Local[7]; | ||
247 : | |||
248 : | for (i = 0; i < 8; i++) | ||
249 : | Local[i] = (Local_Square[i]<<4) - (Local[i]*Local[i]); /* 16*Var(Di) */ | ||
250 : | |||
251 : | Local_Square[0] += (Local_Square[1] + Local_Square[2] + Local_Square[3]); | ||
252 : | Local_Square[4] += (Local_Square[5] + Local_Square[6] + Local_Square[7]); | ||
253 : | |||
254 : | Global_A = (Local_Square[0]<<6) - Global_A*Global_A; /* 64*Var(D) */ | ||
255 : | Global_B = (Local_Square[4]<<6) - Global_B*Global_B; /* 64*Var(D) */ | ||
256 : | |||
257 : | /* Step 3: Calculate contrast masking threshold */ | ||
258 : | { | ||
259 : | Isibaar | 1.2 | uint32_t MASK_A, MASK_B; |
260 : | uint32_t Mult1 = 64, Mult2 = 64; | ||
261 : | |||
262 : | if (Global_A) | ||
263 : | Mult1 = ((Local[0]+Local[1]+Local[2]+Local[3])<<8) / Global_A; | ||
264 : | Isibaar | 1.1 | |
265 : | Isibaar | 1.2 | if (Global_B) |
266 : | Mult2 = ((Local[4]+Local[5]+Local[6]+Local[7])<<8) / Global_B; | ||
267 : | Isibaar | 1.1 | |
268 : | Isibaar | 1.2 | MASK_A = isqrt(2*Sum_A*Mult1) + 16; |
269 : | MASK_B = isqrt(2*Sum_B*Mult2) + 16; | ||
270 : | Isibaar | 1.1 | |
271 : | if (MASK_B > MASK_A) /* MAX(MASK_A, MASK_B) */ | ||
272 : | Isibaar | 1.2 | MASK = (uint32_t) MASK_B; |
273 : | Isibaar | 1.1 | else |
274 : | Isibaar | 1.2 | MASK = (uint32_t) MASK_A; |
275 : | Isibaar | 1.1 | } |
276 : | |||
277 : | /* Step 4: Calculate MSE of DCT coeffs reduced by masking effect */ | ||
278 : | for (j = 0; j < 8; j++) { | ||
279 : | for (i = 0; i < 8; i++) { | ||
280 : | uint32_t Thresh = (MASK * Inv_iMask_Coeff[j*8 + i] + 128) >> 8; | ||
281 : | uint32_t u = abs(DCT_A[j*8 + i] - DCT_B[j*8 + i]) << 10; | ||
282 : | |||
283 : | if ((i|j)>0) { | ||
284 : | if (u < Thresh) | ||
285 : | u = 0; /* The error is not perceivable */ | ||
286 : | else | ||
287 : | u -= Thresh; | ||
288 : | } | ||
289 : | |||
290 : | { | ||
291 : | uint64_t tmp = (u * iCSF_Coeff[j*8 + i] + 512) >> 10; | ||
292 : | MSE_H += (uint32_t) ((tmp * tmp) >> 12); | ||
293 : | } | ||
294 : | } | ||
295 : | } | ||
296 : | return MSE_H; | ||
297 : | } | ||
298 : | |||
299 : | #endif | ||
300 : | |||
301 : | static void psnrhvsm_after(xvid_plg_data_t *data, psnrhvsm_data_t *psnrhvsm) | ||
302 : | { | ||
303 : | DECLARE_ALIGNED_MATRIX(DCT, 2, 64, int16_t, CACHE_LINE); | ||
304 : | Isibaar | 1.2 | uint32_t x, y, stride = data->original.stride[0]; |
305 : | Isibaar | 1.1 | int16_t *DCT_A = &DCT[0], *DCT_B = &DCT[64]; |
306 : | uint8_t *IMG_A = (uint8_t *) data->original.plane[0]; | ||
307 : | uint8_t *IMG_B = (uint8_t *) data->current.plane[0]; | ||
308 : | Isibaar | 1.2 | uint64_t MSE_H = 0; |
309 : | Isibaar | 1.1 | |
310 : | for (y = 0; y < data->height; y += 8) { | ||
311 : | for (x = 0; x < data->width; x += 8) { | ||
312 : | int offset = y*stride + x; | ||
313 : | |||
314 : | emms(); | ||
315 : | |||
316 : | /* Transfer data */ | ||
317 : | transfer_8to16copy(DCT_A, IMG_A + offset, stride); | ||
318 : | transfer_8to16copy(DCT_B, IMG_B + offset, stride); | ||
319 : | |||
320 : | /* Perform DCT */ | ||
321 : | fdct(DCT_A); | ||
322 : | fdct(DCT_B); | ||
323 : | |||
324 : | emms(); | ||
325 : | |||
326 : | /* Calculate MSE_H reduced by contrast masking effect */ | ||
327 : | MSE_H += Calc_MSE_H(DCT_A, DCT_B, IMG_A + offset, IMG_B + offset, stride); | ||
328 : | } | ||
329 : | } | ||
330 : | |||
331 : | Isibaar | 1.2 | x = 4*MSE_H / (data->width * data->height); |
332 : | psnrhvsm->mse_sum += x; | ||
333 : | Isibaar | 1.1 | psnrhvsm->frame_cnt++; |
334 : | |||
335 : | Isibaar | 1.2 | printf(" psnrhvsm: %2.2f\n", sse_to_PSNR(x, 1024)); |
336 : | Isibaar | 1.1 | } |
337 : | |||
338 : | static int psnrhvsm_create(xvid_plg_create_t *create, void **handle) | ||
339 : | { | ||
340 : | psnrhvsm_data_t *psnrhvsm; | ||
341 : | psnrhvsm = (psnrhvsm_data_t *) malloc(sizeof(psnrhvsm_data_t)); | ||
342 : | |||
343 : | psnrhvsm->mse_sum = 0; | ||
344 : | psnrhvsm->frame_cnt = 0; | ||
345 : | |||
346 : | *(handle) = (void*) psnrhvsm; | ||
347 : | return 0; | ||
348 : | } | ||
349 : | |||
350 : | int xvid_plugin_psnrhvsm(void *handle, int opt, void *param1, void *param2) | ||
351 : | { | ||
352 : | switch(opt) { | ||
353 : | case(XVID_PLG_INFO): | ||
354 : | ((xvid_plg_info_t *)param1)->flags = XVID_REQORIGINAL; | ||
355 : | break; | ||
356 : | case(XVID_PLG_CREATE): | ||
357 : | psnrhvsm_create((xvid_plg_create_t *)param1,(void **)param2); | ||
358 : | break; | ||
359 : | case(XVID_PLG_BEFORE): | ||
360 : | case(XVID_PLG_FRAME): | ||
361 : | break; | ||
362 : | case(XVID_PLG_AFTER): | ||
363 : | psnrhvsm_after((xvid_plg_data_t *)param1, (psnrhvsm_data_t *)handle); | ||
364 : | break; | ||
365 : | case(XVID_PLG_DESTROY): | ||
366 : | { | ||
367 : | uint32_t MSE_H; | ||
368 : | psnrhvsm_data_t *psnrhvsm = (psnrhvsm_data_t *)handle; | ||
369 : | |||
370 : | if (psnrhvsm) { | ||
371 : | MSE_H = (uint32_t) (psnrhvsm->mse_sum / psnrhvsm->frame_cnt); | ||
372 : | |||
373 : | emms(); | ||
374 : | Isibaar | 1.2 | printf("Average psnrhvsm: %2.2f\n", sse_to_PSNR(MSE_H, 1024)); |
375 : | Isibaar | 1.1 | free(psnrhvsm); |
376 : | } | ||
377 : | } | ||
378 : | break; | ||
379 : | default: | ||
380 : | break; | ||
381 : | } | ||
382 : | return 0; | ||
383 : | }; |
No admin address has been configured | ViewVC Help |
Powered by ViewVC 1.0.4 |