Libav
|
00001 /* 00002 * DSP functions for Indeo Video Interactive codecs (Indeo4 and Indeo5) 00003 * 00004 * Copyright (c) 2009 Maxim Poliakovski 00005 * 00006 * This file is part of FFmpeg. 00007 * 00008 * FFmpeg is free software; you can redistribute it and/or 00009 * modify it under the terms of the GNU Lesser General Public 00010 * License as published by the Free Software Foundation; either 00011 * version 2.1 of the License, or (at your option) any later version. 00012 * 00013 * FFmpeg is distributed in the hope that it will be useful, 00014 * but WITHOUT ANY WARRANTY; without even the implied warranty of 00015 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU 00016 * Lesser General Public License for more details. 00017 * 00018 * You should have received a copy of the GNU Lesser General Public 00019 * License along with FFmpeg; if not, write to the Free Software 00020 * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 00021 */ 00022 00029 #include "avcodec.h" 00030 #include "dsputil.h" 00031 #include "dwt.h" 00032 #include "ivi_common.h" 00033 #include "ivi_dsp.h" 00034 00035 void ff_ivi_recompose53(const IVIPlaneDesc *plane, uint8_t *dst, 00036 const int dst_pitch, const int num_bands) 00037 { 00038 int x, y, indx; 00039 int32_t p0, p1, p2, p3, tmp0, tmp1, tmp2; 00040 int32_t b0_1, b0_2, b1_1, b1_2, b1_3, b2_1, b2_2, b2_3, b2_4, b2_5, b2_6; 00041 int32_t b3_1, b3_2, b3_3, b3_4, b3_5, b3_6, b3_7, b3_8, b3_9; 00042 int32_t pitch, back_pitch; 00043 const IDWTELEM *b0_ptr, *b1_ptr, *b2_ptr, *b3_ptr; 00044 00045 /* all bands should have the same pitch */ 00046 pitch = plane->bands[0].pitch; 00047 00048 /* pixels at the position "y-1" will be set to pixels at the "y" for the 1st iteration */ 00049 back_pitch = 0; 00050 00051 /* get pointers to the wavelet bands */ 00052 b0_ptr = plane->bands[0].buf; 00053 b1_ptr = plane->bands[1].buf; 00054 b2_ptr = plane->bands[2].buf; 00055 b3_ptr = plane->bands[3].buf; 00056 00057 for (y = 0; y < plane->height; y += 2) { 00058 /* load storage variables with values */ 00059 if (num_bands > 0) { 00060 b0_1 = b0_ptr[0]; 00061 b0_2 = b0_ptr[pitch]; 00062 } 00063 00064 if (num_bands > 1) { 00065 b1_1 = b1_ptr[back_pitch]; 00066 b1_2 = b1_ptr[0]; 00067 b1_3 = b1_1 - b1_2*6 + b1_ptr[pitch]; 00068 } 00069 00070 if (num_bands > 2) { 00071 b2_2 = b2_ptr[0]; // b2[x, y ] 00072 b2_3 = b2_2; // b2[x+1,y ] = b2[x,y] 00073 b2_5 = b2_ptr[pitch]; // b2[x ,y+1] 00074 b2_6 = b2_5; // b2[x+1,y+1] = b2[x,y+1] 00075 } 00076 00077 if (num_bands > 3) { 00078 b3_2 = b3_ptr[back_pitch]; // b3[x ,y-1] 00079 b3_3 = b3_2; // b3[x+1,y-1] = b3[x ,y-1] 00080 b3_5 = b3_ptr[0]; // b3[x ,y ] 00081 b3_6 = b3_5; // b3[x+1,y ] = b3[x ,y ] 00082 b3_8 = b3_2 - b3_5*6 + b3_ptr[pitch]; 00083 b3_9 = b3_8; 00084 } 00085 00086 for (x = 0, indx = 0; x < plane->width; x+=2, indx++) { 00087 /* some values calculated in the previous iterations can */ 00088 /* be reused in the next ones, so do appropriate copying */ 00089 b2_1 = b2_2; // b2[x-1,y ] = b2[x, y ] 00090 b2_2 = b2_3; // b2[x ,y ] = b2[x+1,y ] 00091 b2_4 = b2_5; // b2[x-1,y+1] = b2[x ,y+1] 00092 b2_5 = b2_6; // b2[x ,y+1] = b2[x+1,y+1] 00093 b3_1 = b3_2; // b3[x-1,y-1] = b3[x ,y-1] 00094 b3_2 = b3_3; // b3[x ,y-1] = b3[x+1,y-1] 00095 b3_4 = b3_5; // b3[x-1,y ] = b3[x ,y ] 00096 b3_5 = b3_6; // b3[x ,y ] = b3[x+1,y ] 00097 b3_7 = b3_8; // vert_HPF(x-1) 00098 b3_8 = b3_9; // vert_HPF(x ) 00099 00100 p0 = p1 = p2 = p3 = 0; 00101 00102 /* process the LL-band by applying LPF both vertically and horizontally */ 00103 if (num_bands > 0) { 00104 tmp0 = b0_1; 00105 tmp2 = b0_2; 00106 b0_1 = b0_ptr[indx+1]; 00107 b0_2 = b0_ptr[pitch+indx+1]; 00108 tmp1 = tmp0 + b0_1; 00109 00110 p0 = tmp0 << 4; 00111 p1 = tmp1 << 3; 00112 p2 = (tmp0 + tmp2) << 3; 00113 p3 = (tmp1 + tmp2 + b0_2) << 2; 00114 } 00115 00116 /* process the HL-band by applying HPF vertically and LPF horizontally */ 00117 if (num_bands > 1) { 00118 tmp0 = b1_2; 00119 tmp1 = b1_1; 00120 b1_2 = b1_ptr[indx+1]; 00121 b1_1 = b1_ptr[back_pitch+indx+1]; 00122 00123 tmp2 = tmp1 - tmp0*6 + b1_3; 00124 b1_3 = b1_1 - b1_2*6 + b1_ptr[pitch+indx+1]; 00125 00126 p0 += (tmp0 + tmp1) << 3; 00127 p1 += (tmp0 + tmp1 + b1_1 + b1_2) << 2; 00128 p2 += tmp2 << 2; 00129 p3 += (tmp2 + b1_3) << 1; 00130 } 00131 00132 /* process the LH-band by applying LPF vertically and HPF horizontally */ 00133 if (num_bands > 2) { 00134 b2_3 = b2_ptr[indx+1]; 00135 b2_6 = b2_ptr[pitch+indx+1]; 00136 00137 tmp0 = b2_1 + b2_2; 00138 tmp1 = b2_1 - b2_2*6 + b2_3; 00139 00140 p0 += tmp0 << 3; 00141 p1 += tmp1 << 2; 00142 p2 += (tmp0 + b2_4 + b2_5) << 2; 00143 p3 += (tmp1 + b2_4 - b2_5*6 + b2_6) << 1; 00144 } 00145 00146 /* process the HH-band by applying HPF both vertically and horizontally */ 00147 if (num_bands > 3) { 00148 b3_6 = b3_ptr[indx+1]; // b3[x+1,y ] 00149 b3_3 = b3_ptr[back_pitch+indx+1]; // b3[x+1,y-1] 00150 00151 tmp0 = b3_1 + b3_4; 00152 tmp1 = b3_2 + b3_5; 00153 tmp2 = b3_3 + b3_6; 00154 00155 b3_9 = b3_3 - b3_6*6 + b3_ptr[pitch+indx+1]; 00156 00157 p0 += (tmp0 + tmp1) << 2; 00158 p1 += (tmp0 - tmp1*6 + tmp2) << 1; 00159 p2 += (b3_7 + b3_8) << 1; 00160 p3 += b3_7 - b3_8*6 + b3_9; 00161 } 00162 00163 /* output four pixels */ 00164 dst[x] = av_clip_uint8((p0 >> 6) + 128); 00165 dst[x+1] = av_clip_uint8((p1 >> 6) + 128); 00166 dst[dst_pitch+x] = av_clip_uint8((p2 >> 6) + 128); 00167 dst[dst_pitch+x+1] = av_clip_uint8((p3 >> 6) + 128); 00168 }// for x 00169 00170 dst += dst_pitch << 1; 00171 00172 back_pitch = -pitch; 00173 00174 b0_ptr += pitch; 00175 b1_ptr += pitch; 00176 b2_ptr += pitch; 00177 b3_ptr += pitch; 00178 } 00179 } 00180 00182 #define IVI_SLANT_BFLY(s1, s2, o1, o2, t) \ 00183 t = s1 - s2;\ 00184 o1 = s1 + s2;\ 00185 o2 = t;\ 00186 00187 00188 #define IVI_IREFLECT(s1, s2, o1, o2, t) \ 00189 t = ((s1 + s2*2 + 2) >> 2) + s1;\ 00190 o2 = ((s1*2 - s2 + 2) >> 2) - s2;\ 00191 o1 = t;\ 00192 00193 00194 #define IVI_SLANT_PART4(s1, s2, o1, o2, t) \ 00195 t = s2 + ((s1*4 - s2 + 4) >> 3);\ 00196 o2 = s1 + ((-s1 - s2*4 + 4) >> 3);\ 00197 o1 = t;\ 00198 00199 00200 #define IVI_INV_SLANT8(s1, s4, s8, s5, s2, s6, s3, s7,\ 00201 d1, d2, d3, d4, d5, d6, d7, d8,\ 00202 t0, t1, t2, t3, t4, t5, t6, t7, t8) {\ 00203 IVI_SLANT_PART4(s4, s5, t4, t5, t0);\ 00204 \ 00205 IVI_SLANT_BFLY(s1, t5, t1, t5, t0); IVI_SLANT_BFLY(s2, s6, t2, t6, t0);\ 00206 IVI_SLANT_BFLY(s7, s3, t7, t3, t0); IVI_SLANT_BFLY(t4, s8, t4, t8, t0);\ 00207 \ 00208 IVI_SLANT_BFLY(t1, t2, t1, t2, t0); IVI_IREFLECT (t4, t3, t4, t3, t0);\ 00209 IVI_SLANT_BFLY(t5, t6, t5, t6, t0); IVI_IREFLECT (t8, t7, t8, t7, t0);\ 00210 IVI_SLANT_BFLY(t1, t4, t1, t4, t0); IVI_SLANT_BFLY(t2, t3, t2, t3, t0);\ 00211 IVI_SLANT_BFLY(t5, t8, t5, t8, t0); IVI_SLANT_BFLY(t6, t7, t6, t7, t0);\ 00212 d1 = COMPENSATE(t1);\ 00213 d2 = COMPENSATE(t2);\ 00214 d3 = COMPENSATE(t3);\ 00215 d4 = COMPENSATE(t4);\ 00216 d5 = COMPENSATE(t5);\ 00217 d6 = COMPENSATE(t6);\ 00218 d7 = COMPENSATE(t7);\ 00219 d8 = COMPENSATE(t8);} 00220 00222 #define IVI_INV_SLANT4(s1, s4, s2, s3, d1, d2, d3, d4, t0, t1, t2, t3, t4) {\ 00223 IVI_SLANT_BFLY(s1, s2, t1, t2, t0); IVI_IREFLECT (s4, s3, t4, t3, t0);\ 00224 \ 00225 IVI_SLANT_BFLY(t1, t4, t1, t4, t0); IVI_SLANT_BFLY(t2, t3, t2, t3, t0);\ 00226 d1 = COMPENSATE(t1);\ 00227 d2 = COMPENSATE(t2);\ 00228 d3 = COMPENSATE(t3);\ 00229 d4 = COMPENSATE(t4);} 00230 00231 void ff_ivi_inverse_slant_8x8(const int32_t *in, int16_t *out, uint32_t pitch, const uint8_t *flags) 00232 { 00233 int i; 00234 const int32_t *src; 00235 int32_t *dst; 00236 int tmp[64]; 00237 int t0, t1, t2, t3, t4, t5, t6, t7, t8; 00238 00239 #define COMPENSATE(x) (x) 00240 src = in; 00241 dst = tmp; 00242 for (i = 0; i < 8; i++) { 00243 if (flags[i]) { 00244 IVI_INV_SLANT8(src[0], src[8], src[16], src[24], src[32], src[40], src[48], src[56], 00245 dst[0], dst[8], dst[16], dst[24], dst[32], dst[40], dst[48], dst[56], 00246 t0, t1, t2, t3, t4, t5, t6, t7, t8); 00247 } else 00248 dst[0] = dst[8] = dst[16] = dst[24] = dst[32] = dst[40] = dst[48] = dst[56] = 0; 00249 00250 src++; 00251 dst++; 00252 } 00253 #undef COMPENSATE 00254 00255 #define COMPENSATE(x) ((x + 1)>>1) 00256 src = tmp; 00257 for (i = 0; i < 8; i++) { 00258 if (!src[0] && !src[1] && !src[2] && !src[3] && !src[4] && !src[5] && !src[6] && !src[7]) { 00259 memset(out, 0, 8*sizeof(out[0])); 00260 } else { 00261 IVI_INV_SLANT8(src[0], src[1], src[2], src[3], src[4], src[5], src[6], src[7], 00262 out[0], out[1], out[2], out[3], out[4], out[5], out[6], out[7], 00263 t0, t1, t2, t3, t4, t5, t6, t7, t8); 00264 } 00265 src += 8; 00266 out += pitch; 00267 } 00268 #undef COMPENSATE 00269 } 00270 00271 void ff_ivi_inverse_slant_4x4(const int32_t *in, int16_t *out, uint32_t pitch, const uint8_t *flags) 00272 { 00273 int i; 00274 const int32_t *src; 00275 int32_t *dst; 00276 int tmp[16]; 00277 int t0, t1, t2, t3, t4; 00278 00279 #define COMPENSATE(x) (x) 00280 src = in; 00281 dst = tmp; 00282 for (i = 0; i < 4; i++) { 00283 if (flags[i]) { 00284 IVI_INV_SLANT4(src[0], src[4], src[8], src[12], 00285 dst[0], dst[4], dst[8], dst[12], 00286 t0, t1, t2, t3, t4); 00287 } else 00288 dst[0] = dst[4] = dst[8] = dst[12] = 0; 00289 00290 src++; 00291 dst++; 00292 } 00293 #undef COMPENSATE 00294 00295 #define COMPENSATE(x) ((x + 1)>>1) 00296 src = tmp; 00297 for (i = 0; i < 4; i++) { 00298 if (!src[0] && !src[1] && !src[2] && !src[3]) { 00299 out[0] = out[1] = out[2] = out[3] = 0; 00300 } else { 00301 IVI_INV_SLANT4(src[0], src[1], src[2], src[3], 00302 out[0], out[1], out[2], out[3], 00303 t0, t1, t2, t3, t4); 00304 } 00305 src += 4; 00306 out += pitch; 00307 } 00308 #undef COMPENSATE 00309 } 00310 00311 void ff_ivi_dc_slant_2d(const int32_t *in, int16_t *out, uint32_t pitch, int blk_size) 00312 { 00313 int x, y; 00314 int16_t dc_coeff; 00315 00316 dc_coeff = (*in + 1) >> 1; 00317 00318 for (y = 0; y < blk_size; out += pitch, y++) { 00319 for (x = 0; x < blk_size; x++) 00320 out[x] = dc_coeff; 00321 } 00322 } 00323 00324 void ff_ivi_row_slant8(const int32_t *in, int16_t *out, uint32_t pitch, const uint8_t *flags) 00325 { 00326 int i; 00327 int t0, t1, t2, t3, t4, t5, t6, t7, t8; 00328 00329 #define COMPENSATE(x) ((x + 1)>>1) 00330 for (i = 0; i < 8; i++) { 00331 if (!in[0] && !in[1] && !in[2] && !in[3] && !in[4] && !in[5] && !in[6] && !in[7]) { 00332 memset(out, 0, 8*sizeof(out[0])); 00333 } else { 00334 IVI_INV_SLANT8( in[0], in[1], in[2], in[3], in[4], in[5], in[6], in[7], 00335 out[0], out[1], out[2], out[3], out[4], out[5], out[6], out[7], 00336 t0, t1, t2, t3, t4, t5, t6, t7, t8); 00337 } 00338 in += 8; 00339 out += pitch; 00340 } 00341 #undef COMPENSATE 00342 } 00343 00344 void ff_ivi_dc_row_slant(const int32_t *in, int16_t *out, uint32_t pitch, int blk_size) 00345 { 00346 int x, y; 00347 int16_t dc_coeff; 00348 00349 dc_coeff = (*in + 1) >> 1; 00350 00351 for (x = 0; x < blk_size; x++) 00352 out[x] = dc_coeff; 00353 00354 out += pitch; 00355 00356 for (y = 1; y < blk_size; out += pitch, y++) { 00357 for (x = 0; x < blk_size; x++) 00358 out[x] = 0; 00359 } 00360 } 00361 00362 void ff_ivi_col_slant8(const int32_t *in, int16_t *out, uint32_t pitch, const uint8_t *flags) 00363 { 00364 int i, row2, row4, row8; 00365 int t0, t1, t2, t3, t4, t5, t6, t7, t8; 00366 00367 row2 = pitch << 1; 00368 row4 = pitch << 2; 00369 row8 = pitch << 3; 00370 00371 #define COMPENSATE(x) ((x + 1)>>1) 00372 for (i = 0; i < 8; i++) { 00373 if (flags[i]) { 00374 IVI_INV_SLANT8(in[0], in[8], in[16], in[24], in[32], in[40], in[48], in[56], 00375 out[0], out[pitch], out[row2], out[row2 + pitch], out[row4], 00376 out[row4 + pitch], out[row4 + row2], out[row8 - pitch], 00377 t0, t1, t2, t3, t4, t5, t6, t7, t8); 00378 } else { 00379 out[0] = out[pitch] = out[row2] = out[row2 + pitch] = out[row4] = 00380 out[row4 + pitch] = out[row4 + row2] = out[row8 - pitch] = 0; 00381 } 00382 00383 in++; 00384 out++; 00385 } 00386 #undef COMPENSATE 00387 } 00388 00389 void ff_ivi_dc_col_slant(const int32_t *in, int16_t *out, uint32_t pitch, int blk_size) 00390 { 00391 int x, y; 00392 int16_t dc_coeff; 00393 00394 dc_coeff = (*in + 1) >> 1; 00395 00396 for (y = 0; y < blk_size; out += pitch, y++) { 00397 out[0] = dc_coeff; 00398 for (x = 1; x < blk_size; x++) 00399 out[x] = 0; 00400 } 00401 } 00402 00403 void ff_ivi_put_pixels_8x8(const int32_t *in, int16_t *out, uint32_t pitch, 00404 const uint8_t *flags) 00405 { 00406 int x, y; 00407 00408 for (y = 0; y < 8; out += pitch, in += 8, y++) 00409 for (x = 0; x < 8; x++) 00410 out[x] = in[x]; 00411 } 00412 00413 void ff_ivi_put_dc_pixel_8x8(const int32_t *in, int16_t *out, uint32_t pitch, 00414 int blk_size) 00415 { 00416 int y; 00417 00418 out[0] = in[0]; 00419 memset(out + 1, 0, 7*sizeof(out[0])); 00420 out += pitch; 00421 00422 for (y = 1; y < 8; out += pitch, y++) 00423 memset(out, 0, 8*sizeof(out[0])); 00424 } 00425 00426 #define IVI_MC_TEMPLATE(size, suffix, OP) \ 00427 void ff_ivi_mc_ ## size ##x## size ## suffix (int16_t *buf, const int16_t *ref_buf, \ 00428 uint32_t pitch, int mc_type) \ 00429 { \ 00430 int i, j; \ 00431 const int16_t *wptr; \ 00432 \ 00433 switch (mc_type) { \ 00434 case 0: /* fullpel (no interpolation) */ \ 00435 for (i = 0; i < size; i++, buf += pitch, ref_buf += pitch) { \ 00436 for (j = 0; j < size; j++) {\ 00437 OP(buf[j], ref_buf[j]); \ 00438 } \ 00439 } \ 00440 break; \ 00441 case 1: /* horizontal halfpel interpolation */ \ 00442 for (i = 0; i < size; i++, buf += pitch, ref_buf += pitch) \ 00443 for (j = 0; j < size; j++) \ 00444 OP(buf[j], (ref_buf[j] + ref_buf[j+1]) >> 1); \ 00445 break; \ 00446 case 2: /* vertical halfpel interpolation */ \ 00447 wptr = ref_buf + pitch; \ 00448 for (i = 0; i < size; i++, buf += pitch, wptr += pitch, ref_buf += pitch) \ 00449 for (j = 0; j < size; j++) \ 00450 OP(buf[j], (ref_buf[j] + wptr[j]) >> 1); \ 00451 break; \ 00452 case 3: /* vertical and horizontal halfpel interpolation */ \ 00453 wptr = ref_buf + pitch; \ 00454 for (i = 0; i < size; i++, buf += pitch, wptr += pitch, ref_buf += pitch) \ 00455 for (j = 0; j < size; j++) \ 00456 OP(buf[j], (ref_buf[j] + ref_buf[j+1] + wptr[j] + wptr[j+1]) >> 2); \ 00457 break; \ 00458 } \ 00459 } \ 00460 00461 #define OP_PUT(a, b) (a) = (b) 00462 #define OP_ADD(a, b) (a) += (b) 00463 00464 IVI_MC_TEMPLATE(8, _no_delta, OP_PUT); 00465 IVI_MC_TEMPLATE(8, _delta, OP_ADD); 00466 IVI_MC_TEMPLATE(4, _no_delta, OP_PUT); 00467 IVI_MC_TEMPLATE(4, _delta, OP_ADD);