/* * Copyright (c) 2002-2007, Communications and Remote Sensing Laboratory, Universite catholique de Louvain (UCL), Belgium * Copyright (c) 2002-2007, Professor Benoit Macq * Copyright (c) 2001-2003, David Janssens * Copyright (c) 2002-2003, Yannick Verschueren * Copyright (c) 2003-2007, Francois-Olivier Devaux and Antonin Descampe * Copyright (c) 2005, Herve Drolon, FreeImage Team * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS `AS IS' * AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ #ifdef __SSE__ #include <xmmintrin.h> #endif #include "opj_includes.h" /* <summary> */ /* This table contains the norms of the basis function of the reversible MCT. */ /* </summary> */ static const double mct_norms[3] = { 1.732, .8292, .8292 }; /* <summary> */ /* This table contains the norms of the basis function of the irreversible MCT. */ /* </summary> */ static const double mct_norms_real[3] = { 1.732, 1.805, 1.573 }; const OPJ_FLOAT64 * get_mct_norms () { return mct_norms; } const OPJ_FLOAT64 * get_mct_norms_real () { return mct_norms_real; } /* <summary> */ /* Foward reversible MCT. */ /* </summary> */ void mct_encode( int* restrict c0, int* restrict c1, int* restrict c2, int n) { int i; for(i = 0; i < n; ++i) { int r = c0[i]; int g = c1[i]; int b = c2[i]; int y = (r + (g * 2) + b) >> 2; int u = b - g; int v = r - g; c0[i] = y; c1[i] = u; c2[i] = v; } } /* <summary> */ /* Inverse reversible MCT. */ /* </summary> */ void mct_decode( int* restrict c0, int* restrict c1, int* restrict c2, int n) { int i; for (i = 0; i < n; ++i) { int y = c0[i]; int u = c1[i]; int v = c2[i]; int g = y - ((u + v) >> 2); int r = v + g; int b = u + g; c0[i] = r; c1[i] = g; c2[i] = b; } } /* <summary> */ /* Get norm of basis function of reversible MCT. */ /* </summary> */ double mct_getnorm(int compno) { return mct_norms[compno]; } /* <summary> */ /* Foward irreversible MCT. */ /* </summary> */ void mct_encode_real( int* restrict c0, int* restrict c1, int* restrict c2, int n) { int i; for(i = 0; i < n; ++i) { int r = c0[i]; int g = c1[i]; int b = c2[i]; int y = fix_mul(r, 2449) + fix_mul(g, 4809) + fix_mul(b, 934); int u = -fix_mul(r, 1382) - fix_mul(g, 2714) + fix_mul(b, 4096); int v = fix_mul(r, 4096) - fix_mul(g, 3430) - fix_mul(b, 666); c0[i] = y; c1[i] = u; c2[i] = v; } } /* <summary> */ /* Inverse irreversible MCT. */ /* </summary> */ void mct_decode_real( float* restrict c0, float* restrict c1, float* restrict c2, int n) { int i; #ifdef __SSE__ __m128 vrv, vgu, vgv, vbu; vrv = _mm_set1_ps(1.402f); vgu = _mm_set1_ps(0.34413f); vgv = _mm_set1_ps(0.71414f); vbu = _mm_set1_ps(1.772f); for (i = 0; i < (n >> 3); ++i) { __m128 vy, vu, vv; __m128 vr, vg, vb; vy = _mm_load_ps(c0); vu = _mm_load_ps(c1); vv = _mm_load_ps(c2); vr = _mm_add_ps(vy, _mm_mul_ps(vv, vrv)); vg = _mm_sub_ps(_mm_sub_ps(vy, _mm_mul_ps(vu, vgu)), _mm_mul_ps(vv, vgv)); vb = _mm_add_ps(vy, _mm_mul_ps(vu, vbu)); _mm_store_ps(c0, vr); _mm_store_ps(c1, vg); _mm_store_ps(c2, vb); c0 += 4; c1 += 4; c2 += 4; vy = _mm_load_ps(c0); vu = _mm_load_ps(c1); vv = _mm_load_ps(c2); vr = _mm_add_ps(vy, _mm_mul_ps(vv, vrv)); vg = _mm_sub_ps(_mm_sub_ps(vy, _mm_mul_ps(vu, vgu)), _mm_mul_ps(vv, vgv)); vb = _mm_add_ps(vy, _mm_mul_ps(vu, vbu)); _mm_store_ps(c0, vr); _mm_store_ps(c1, vg); _mm_store_ps(c2, vb); c0 += 4; c1 += 4; c2 += 4; } n &= 7; #endif for(i = 0; i < n; ++i) { float y = c0[i]; float u = c1[i]; float v = c2[i]; float r = y + (v * 1.402f); float g = y - (u * 0.34413f) - (v * (0.71414f)); float b = y + (u * 1.772f); c0[i] = r; c1[i] = g; c2[i] = b; } } /* <summary> */ /* Get norm of basis function of irreversible MCT. */ /* </summary> */ double mct_getnorm_real(int compno) { return mct_norms_real[compno]; } opj_bool mct_encode_custom( /* MCT data */ OPJ_BYTE * pCodingdata, /* size of components */ OPJ_UINT32 n, /* components */ OPJ_BYTE ** pData, /* nb of components (i.e. size of pData) */ OPJ_UINT32 pNbComp, /* tells if the data is signed */ OPJ_UINT32 isSigned) { OPJ_FLOAT32 * lMct = (OPJ_FLOAT32 *) pCodingdata; OPJ_UINT32 i; OPJ_UINT32 j; OPJ_UINT32 k; OPJ_UINT32 lNbMatCoeff = pNbComp * pNbComp; OPJ_INT32 * lCurrentData = 00; OPJ_INT32 * lCurrentMatrix = 00; OPJ_INT32 ** lData = (OPJ_INT32 **) pData; OPJ_UINT32 lMultiplicator = 1 << 13; OPJ_INT32 * lMctPtr; lCurrentData = (OPJ_INT32 *) opj_malloc((pNbComp + lNbMatCoeff) * sizeof(OPJ_INT32)); if (! lCurrentData) { return OPJ_FALSE; } lCurrentMatrix = lCurrentData + pNbComp; for (i =0;i<lNbMatCoeff;++i) { lCurrentMatrix[i] = (OPJ_INT32) (*(lMct++) * lMultiplicator); } for (i = 0; i < n; ++i) { lMctPtr = lCurrentMatrix; for (j=0;j<pNbComp;++j) { lCurrentData[j] = (*(lData[j])); } for (j=0;j<pNbComp;++j) { *(lData[j]) = 0; for (k=0;k<pNbComp;++k) { *(lData[j]) += fix_mul(*lMctPtr, lCurrentData[k]); ++lMctPtr; } ++lData[j]; } } opj_free(lCurrentData); return OPJ_TRUE; } opj_bool mct_decode_custom( /* MCT data */ OPJ_BYTE * pDecodingData, /* size of components */ OPJ_UINT32 n, /* components */ OPJ_BYTE ** pData, /* nb of components (i.e. size of pData) */ OPJ_UINT32 pNbComp, /* tells if the data is signed */ OPJ_UINT32 isSigned) { OPJ_FLOAT32 * lMct; OPJ_UINT32 i; OPJ_UINT32 j; OPJ_UINT32 k; OPJ_FLOAT32 * lCurrentData = 00; OPJ_FLOAT32 * lCurrentResult = 00; OPJ_FLOAT32 ** lData = (OPJ_FLOAT32 **) pData; lCurrentData = (OPJ_FLOAT32 *) opj_malloc (2 * pNbComp * sizeof(OPJ_FLOAT32)); if (! lCurrentData) { return OPJ_FALSE; } lCurrentResult = lCurrentData + pNbComp; for (i = 0; i < n; ++i) { lMct = (OPJ_FLOAT32 *) pDecodingData; for (j=0;j<pNbComp;++j) { lCurrentData[j] = (OPJ_FLOAT32) (*(lData[j])); } for (j=0;j<pNbComp;++j) { lCurrentResult[j] = 0; for (k=0;k<pNbComp;++k) { lCurrentResult[j] += *(lMct++) * lCurrentData[k]; } *(lData[j]++) = (OPJ_FLOAT32) (lCurrentResult[j]); } } opj_free(lCurrentData); return OPJ_TRUE; } void opj_calculate_norms( OPJ_FLOAT64 * pNorms, OPJ_UINT32 pNbComps, OPJ_FLOAT32 * pMatrix) { OPJ_UINT32 i,j,lIndex; OPJ_FLOAT32 lCurrentValue; OPJ_FLOAT64 * lNorms = (OPJ_FLOAT64 *) pNorms; OPJ_FLOAT32 * lMatrix = (OPJ_FLOAT32 *) pMatrix; for (i=0;i<pNbComps;++i) { lNorms[i] = 0; lIndex = i; for (j=0;j<pNbComps;++j) { lCurrentValue = lMatrix[lIndex]; lIndex += pNbComps; lNorms[i] += lCurrentValue * lCurrentValue; } lNorms[i] = sqrt(lNorms[i]); } }