/*M/////////////////////////////////////////////////////////////////////////////////////// // // IMPORTANT: READ BEFORE DOWNLOADING, COPYING, INSTALLING OR USING. // // By downloading, copying, installing or using the software you agree to this license. // If you do not agree to this license, do not download, install, // copy or use the software. // // // Intel License Agreement // For Open Source Computer Vision Library // // Copyright (C) 2000, Intel Corporation, all rights reserved. // Third party copyrights are property of their respective owners. // // Redistribution and use in source and binary forms, with or without modification, // are permitted provided that the following conditions are met: // // * Redistribution's of source code must retain the above copyright notice, // this list of conditions and the following disclaimer. // // * Redistribution's 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. // // * The name of Intel Corporation may not be used to endorse or promote products // derived from this software without specific prior written permission. // // 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 Intel Corporation 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. // //M*/ #include "_cxcore.h" /****************************************************************************************\ * [scaled] Identity matrix initialization * \****************************************************************************************/ CV_IMPL void cvSetIdentity( CvArr* array, CvScalar value ) { CV_FUNCNAME( "cvSetIdentity" ); __BEGIN__; CvMat stub, *mat = (CvMat*)array; CvSize size; int i, k, len, step; int type, pix_size; uchar* data = 0; double buf[4]; if( !CV_IS_MAT( mat )) { int coi = 0; CV_CALL( mat = cvGetMat( mat, &stub, &coi )); if( coi != 0 ) CV_ERROR( CV_BadCOI, "coi is not supported" ); } size = cvGetMatSize( mat ); len = CV_IMIN( size.width, size.height ); type = CV_MAT_TYPE(mat->type); pix_size = CV_ELEM_SIZE(type); size.width *= pix_size; if( CV_IS_MAT_CONT( mat->type )) { size.width *= size.height; size.height = 1; } data = mat->data.ptr; step = mat->step; if( step == 0 ) step = CV_STUB_STEP; IPPI_CALL( icvSetZero_8u_C1R( data, step, size )); step += pix_size; if( type == CV_32FC1 ) { float val = (float)value.val[0]; float* _data = (float*)data; step /= sizeof(_data[0]); len *= step; for( i = 0; i < len; i += step ) _data[i] = val; } else if( type == CV_64FC1 ) { double val = value.val[0]; double* _data = (double*)data; step /= sizeof(_data[0]); len *= step; for( i = 0; i < len; i += step ) _data[i] = val; } else { uchar* val_ptr = (uchar*)buf; cvScalarToRawData( &value, buf, type, 0 ); len *= step; for( i = 0; i < len; i += step ) for( k = 0; k < pix_size; k++ ) data[i+k] = val_ptr[k]; } __END__; } /****************************************************************************************\ * Trace of the matrix * \****************************************************************************************/ CV_IMPL CvScalar cvTrace( const CvArr* array ) { CvScalar sum = {{0,0,0,0}}; CV_FUNCNAME( "cvTrace" ); __BEGIN__; CvMat stub, *mat = 0; if( CV_IS_MAT( array )) { mat = (CvMat*)array; int type = CV_MAT_TYPE(mat->type); int size = MIN(mat->rows,mat->cols); uchar* data = mat->data.ptr; if( type == CV_32FC1 ) { int step = mat->step + sizeof(float); for( ; size--; data += step ) sum.val[0] += *(float*)data; EXIT; } if( type == CV_64FC1 ) { int step = mat->step + sizeof(double); for( ; size--; data += step ) sum.val[0] += *(double*)data; EXIT; } } CV_CALL( mat = cvGetDiag( array, &stub )); CV_CALL( sum = cvSum( mat )); __END__; return sum; } /****************************************************************************************\ * Matrix transpose * \****************************************************************************************/ /////////////////// macros for inplace transposition of square matrix //////////////////// #define ICV_DEF_TRANSP_INP_CASE_C1( \ arrtype, len ) \ { \ arrtype* arr1 = arr; \ step /= sizeof(arr[0]); \ \ while( --len ) \ { \ arr += step, arr1++; \ arrtype* arr2 = arr; \ arrtype* arr3 = arr1; \ \ do \ { \ arrtype t0 = arr2[0]; \ arrtype t1 = arr3[0]; \ arr2[0] = t1; \ arr3[0] = t0; \ \ arr2++; \ arr3 += step; \ } \ while( arr2 != arr3 ); \ } \ } #define ICV_DEF_TRANSP_INP_CASE_C3( \ arrtype, len ) \ { \ arrtype* arr1 = arr; \ int y; \ step /= sizeof(arr[0]); \ \ for( y = 1; y < len; y++ ) \ { \ arr += step, arr1 += 3; \ arrtype* arr2 = arr; \ arrtype* arr3 = arr1; \ \ for( ; arr2!=arr3; arr2+=3, \ arr3+=step )\ { \ arrtype t0 = arr2[0]; \ arrtype t1 = arr3[0]; \ arr2[0] = t1; \ arr3[0] = t0; \ t0 = arr2[1]; \ t1 = arr3[1]; \ arr2[1] = t1; \ arr3[1] = t0; \ t0 = arr2[2]; \ t1 = arr3[2]; \ arr2[2] = t1; \ arr3[2] = t0; \ } \ } \ } #define ICV_DEF_TRANSP_INP_CASE_C4( \ arrtype, len ) \ { \ arrtype* arr1 = arr; \ int y; \ step /= sizeof(arr[0]); \ \ for( y = 1; y < len; y++ ) \ { \ arr += step, arr1 += 4; \ arrtype* arr2 = arr; \ arrtype* arr3 = arr1; \ \ for( ; arr2!=arr3; arr2+=4, \ arr3+=step )\ { \ arrtype t0 = arr2[0]; \ arrtype t1 = arr3[0]; \ arr2[0] = t1; \ arr3[0] = t0; \ t0 = arr2[1]; \ t1 = arr3[1]; \ arr2[1] = t1; \ arr3[1] = t0; \ t0 = arr2[2]; \ t1 = arr3[2]; \ arr2[2] = t1; \ arr3[2] = t0; \ t0 = arr2[3]; \ t1 = arr3[3]; \ arr2[3] = t1; \ arr3[3] = t0; \ } \ } \ } //////////////// macros for non-inplace transposition of rectangular matrix ////////////// #define ICV_DEF_TRANSP_CASE_C1( arrtype ) \ { \ int x, y; \ srcstep /= sizeof(src[0]); \ dststep /= sizeof(dst[0]); \ \ for( y = 0; y <= size.height - 2; y += 2, \ src += 2*srcstep, dst += 2 ) \ { \ const arrtype* src1 = src + srcstep; \ arrtype* dst1 = dst; \ \ for( x = 0; x <= size.width - 2; \ x += 2, dst1 += dststep ) \ { \ arrtype t0 = src[x]; \ arrtype t1 = src1[x]; \ dst1[0] = t0; \ dst1[1] = t1; \ dst1 += dststep; \ \ t0 = src[x + 1]; \ t1 = src1[x + 1]; \ dst1[0] = t0; \ dst1[1] = t1; \ } \ \ if( x < size.width ) \ { \ arrtype t0 = src[x]; \ arrtype t1 = src1[x]; \ dst1[0] = t0; \ dst1[1] = t1; \ } \ } \ \ if( y < size.height ) \ { \ arrtype* dst1 = dst; \ for( x = 0; x <= size.width - 2; \ x += 2, dst1 += 2*dststep ) \ { \ arrtype t0 = src[x]; \ arrtype t1 = src[x + 1]; \ dst1[0] = t0; \ dst1[dststep] = t1; \ } \ \ if( x < size.width ) \ { \ arrtype t0 = src[x]; \ dst1[0] = t0; \ } \ } \ } #define ICV_DEF_TRANSP_CASE_C3( arrtype ) \ { \ size.width *= 3; \ srcstep /= sizeof(src[0]); \ dststep /= sizeof(dst[0]); \ \ for( ; size.height--; src+=srcstep, dst+=3 )\ { \ int x; \ arrtype* dst1 = dst; \ \ for( x = 0; x < size.width; x += 3, \ dst1 += dststep ) \ { \ arrtype t0 = src[x]; \ arrtype t1 = src[x + 1]; \ arrtype t2 = src[x + 2]; \ \ dst1[0] = t0; \ dst1[1] = t1; \ dst1[2] = t2; \ } \ } \ } #define ICV_DEF_TRANSP_CASE_C4( arrtype ) \ { \ size.width *= 4; \ srcstep /= sizeof(src[0]); \ dststep /= sizeof(dst[0]); \ \ for( ; size.height--; src+=srcstep, dst+=4 )\ { \ int x; \ arrtype* dst1 = dst; \ \ for( x = 0; x < size.width; x += 4, \ dst1 += dststep ) \ { \ arrtype t0 = src[x]; \ arrtype t1 = src[x + 1]; \ \ dst1[0] = t0; \ dst1[1] = t1; \ \ t0 = src[x + 2]; \ t1 = src[x + 3]; \ \ dst1[2] = t0; \ dst1[3] = t1; \ } \ } \ } #define ICV_DEF_TRANSP_INP_FUNC( flavor, arrtype, cn ) \ static CvStatus CV_STDCALL \ icvTranspose_##flavor( arrtype* arr, int step, CvSize size )\ { \ assert( size.width == size.height ); \ \ ICV_DEF_TRANSP_INP_CASE_C##cn( arrtype, size.width ) \ return CV_OK; \ } #define ICV_DEF_TRANSP_FUNC( flavor, arrtype, cn ) \ static CvStatus CV_STDCALL \ icvTranspose_##flavor( const arrtype* src, int srcstep, \ arrtype* dst, int dststep, CvSize size )\ { \ ICV_DEF_TRANSP_CASE_C##cn( arrtype ) \ return CV_OK; \ } ICV_DEF_TRANSP_INP_FUNC( 8u_C1IR, uchar, 1 ) ICV_DEF_TRANSP_INP_FUNC( 8u_C2IR, ushort, 1 ) ICV_DEF_TRANSP_INP_FUNC( 8u_C3IR, uchar, 3 ) ICV_DEF_TRANSP_INP_FUNC( 16u_C2IR, int, 1 ) ICV_DEF_TRANSP_INP_FUNC( 16u_C3IR, ushort, 3 ) ICV_DEF_TRANSP_INP_FUNC( 32s_C2IR, int64, 1 ) ICV_DEF_TRANSP_INP_FUNC( 32s_C3IR, int, 3 ) ICV_DEF_TRANSP_INP_FUNC( 64s_C2IR, int, 4 ) ICV_DEF_TRANSP_INP_FUNC( 64s_C3IR, int64, 3 ) ICV_DEF_TRANSP_INP_FUNC( 64s_C4IR, int64, 4 ) ICV_DEF_TRANSP_FUNC( 8u_C1R, uchar, 1 ) ICV_DEF_TRANSP_FUNC( 8u_C2R, ushort, 1 ) ICV_DEF_TRANSP_FUNC( 8u_C3R, uchar, 3 ) ICV_DEF_TRANSP_FUNC( 16u_C2R, int, 1 ) ICV_DEF_TRANSP_FUNC( 16u_C3R, ushort, 3 ) ICV_DEF_TRANSP_FUNC( 32s_C2R, int64, 1 ) ICV_DEF_TRANSP_FUNC( 32s_C3R, int, 3 ) ICV_DEF_TRANSP_FUNC( 64s_C2R, int, 4 ) ICV_DEF_TRANSP_FUNC( 64s_C3R, int64, 3 ) ICV_DEF_TRANSP_FUNC( 64s_C4R, int64, 4 ) CV_DEF_INIT_PIXSIZE_TAB_2D( Transpose, R ) CV_DEF_INIT_PIXSIZE_TAB_2D( Transpose, IR ) CV_IMPL void cvTranspose( const CvArr* srcarr, CvArr* dstarr ) { static CvBtFuncTable tab, inp_tab; static int inittab = 0; CV_FUNCNAME( "cvTranspose" ); __BEGIN__; CvMat sstub, *src = (CvMat*)srcarr; CvMat dstub, *dst = (CvMat*)dstarr; CvSize size; int type, pix_size; if( !inittab ) { icvInitTransposeIRTable( &inp_tab ); icvInitTransposeRTable( &tab ); inittab = 1; } if( !CV_IS_MAT( src )) { int coi = 0; CV_CALL( src = cvGetMat( src, &sstub, &coi )); if( coi != 0 ) CV_ERROR( CV_BadCOI, "coi is not supported" ); } type = CV_MAT_TYPE( src->type ); pix_size = CV_ELEM_SIZE(type); size = cvGetMatSize( src ); if( dstarr == srcarr ) { dst = src; } else { if( !CV_IS_MAT( dst )) { int coi = 0; CV_CALL( dst = cvGetMat( dst, &dstub, &coi )); if( coi != 0 ) CV_ERROR( CV_BadCOI, "coi is not supported" ); } if( !CV_ARE_TYPES_EQ( src, dst )) CV_ERROR( CV_StsUnmatchedFormats, "" ); if( size.width != dst->height || size.height != dst->width ) CV_ERROR( CV_StsUnmatchedSizes, "" ); } if( src->data.ptr == dst->data.ptr ) { if( size.width == size.height ) { CvFunc2D_1A func = (CvFunc2D_1A)(inp_tab.fn_2d[pix_size]); if( !func ) CV_ERROR( CV_StsUnsupportedFormat, "" ); IPPI_CALL( func( src->data.ptr, src->step, size )); } else { if( size.width != 1 && size.height != 1 ) CV_ERROR( CV_StsBadSize, "Rectangular matrix can not be transposed inplace" ); if( !CV_IS_MAT_CONT( src->type & dst->type )) CV_ERROR( CV_StsBadFlag, "In case of inplace column/row transposition " "both source and destination must be continuous" ); if( dst == src ) { int t; CV_SWAP( dst->width, dst->height, t ); dst->step = dst->height == 1 ? 0 : pix_size; } } } else { CvFunc2D_2A func = (CvFunc2D_2A)(tab.fn_2d[pix_size]); if( !func ) CV_ERROR( CV_StsUnsupportedFormat, "" ); IPPI_CALL( func( src->data.ptr, src->step, dst->data.ptr, dst->step, size )); } __END__; } /****************************************************************************************\ * LU decomposition/back substitution * \****************************************************************************************/ CV_IMPL void cvCompleteSymm( CvMat* matrix, int LtoR ) { CV_FUNCNAME( "cvCompleteSymm" ); __BEGIN__; int i, j, nrows; CV_ASSERT( CV_IS_MAT(matrix) && matrix->rows == matrix->cols ); nrows = matrix->rows; if( CV_MAT_TYPE(matrix->type) == CV_32FC1 || CV_MAT_TYPE(matrix->type) == CV_32SC1 ) { int* data = matrix->data.i; int step = matrix->step/sizeof(data[0]); int j0 = 0, j1 = nrows; for( i = 0; i < nrows; i++ ) { if( !LtoR ) j1 = i; else j0 = i+1; for( j = j0; j < j1; j++ ) data[i*step + j] = data[j*step + i]; } } else if( CV_MAT_TYPE(matrix->type) == CV_64FC1 ) { double* data = matrix->data.db; int step = matrix->step/sizeof(data[0]); int j0 = 0, j1 = nrows; for( i = 0; i < nrows; i++ ) { if( !LtoR ) j1 = i; else j0 = i+1; for( j = j0; j < j1; j++ ) data[i*step + j] = data[j*step + i]; } } else CV_ERROR( CV_StsUnsupportedFormat, "" ); __END__; } /****************************************************************************************\ * LU decomposition/back substitution * \****************************************************************************************/ #define arrtype float #define temptype double typedef CvStatus (CV_STDCALL * CvLUDecompFunc)( double* A, int stepA, CvSize sizeA, void* B, int stepB, CvSize sizeB, double* det ); typedef CvStatus (CV_STDCALL * CvLUBackFunc)( double* A, int stepA, CvSize sizeA, void* B, int stepB, CvSize sizeB ); #define ICV_DEF_LU_DECOMP_FUNC( flavor, arrtype ) \ static CvStatus CV_STDCALL \ icvLUDecomp_##flavor( double* A, int stepA, CvSize sizeA, \ arrtype* B, int stepB, CvSize sizeB, double* _det ) \ { \ int n = sizeA.width; \ int m = 0, i; \ double det = 1; \ \ assert( sizeA.width == sizeA.height ); \ \ if( B ) \ { \ assert( sizeA.height == sizeB.height ); \ m = sizeB.width; \ } \ stepA /= sizeof(A[0]); \ stepB /= sizeof(B[0]); \ \ for( i = 0; i < n; i++, A += stepA, B += stepB ) \ { \ int j, k = i; \ double* tA = A; \ arrtype* tB = 0; \ double kval = fabs(A[i]), tval; \ \ /* find the pivot element */ \ for( j = i + 1; j < n; j++ ) \ { \ tA += stepA; \ tval = fabs(tA[i]); \ \ if( tval > kval ) \ { \ kval = tval; \ k = j; \ } \ } \ \ if( kval == 0 ) \ { \ det = 0; \ break; \ } \ \ /* swap rows */ \ if( k != i ) \ { \ tA = A + stepA*(k - i); \ det = -det; \ \ for( j = i; j < n; j++ ) \ { \ double t; \ CV_SWAP( A[j], tA[j], t ); \ } \ \ if( m > 0 ) \ { \ tB = B + stepB*(k - i); \ \ for( j = 0; j < m; j++ ) \ { \ arrtype t = B[j]; \ CV_SWAP( B[j], tB[j], t ); \ } \ } \ } \ \ tval = 1./A[i]; \ det *= A[i]; \ tA = A; \ tB = B; \ A[i] = tval; /* to replace division with multiplication in LUBack */ \ \ /* update matrix and the right side of the system */ \ for( j = i + 1; j < n; j++ ) \ { \ tA += stepA; \ tB += stepB; \ double alpha = -tA[i]*tval; \ \ for( k = i + 1; k < n; k++ ) \ tA[k] = tA[k] + alpha*A[k]; \ \ if( m > 0 ) \ for( k = 0; k < m; k++ ) \ tB[k] = (arrtype)(tB[k] + alpha*B[k]); \ } \ } \ \ if( _det ) \ *_det = det; \ \ return CV_OK; \ } ICV_DEF_LU_DECOMP_FUNC( 32f, float ) ICV_DEF_LU_DECOMP_FUNC( 64f, double ) #define ICV_DEF_LU_BACK_FUNC( flavor, arrtype ) \ static CvStatus CV_STDCALL \ icvLUBack_##flavor( double* A, int stepA, CvSize sizeA, \ arrtype* B, int stepB, CvSize sizeB ) \ { \ int n = sizeA.width; \ int m = sizeB.width, i; \ \ assert( m > 0 && sizeA.width == sizeA.height && \ sizeA.height == sizeB.height ); \ stepA /= sizeof(A[0]); \ stepB /= sizeof(B[0]); \ \ A += stepA*(n - 1); \ B += stepB*(n - 1); \ \ for( i = n - 1; i >= 0; i--, A -= stepA ) \ { \ int j, k; \ for( j = 0; j < m; j++ ) \ { \ arrtype* tB = B + j; \ double x = 0; \ \ for( k = n - 1; k > i; k--, tB -= stepB ) \ x += A[k]*tB[0]; \ \ tB[0] = (arrtype)((tB[0] - x)*A[i]); \ } \ } \ \ return CV_OK; \ } ICV_DEF_LU_BACK_FUNC( 32f, float ) ICV_DEF_LU_BACK_FUNC( 64f, double ) static CvFuncTable lu_decomp_tab, lu_back_tab; static int lu_inittab = 0; static void icvInitLUTable( CvFuncTable* decomp_tab, CvFuncTable* back_tab ) { decomp_tab->fn_2d[0] = (void*)icvLUDecomp_32f; decomp_tab->fn_2d[1] = (void*)icvLUDecomp_64f; back_tab->fn_2d[0] = (void*)icvLUBack_32f; back_tab->fn_2d[1] = (void*)icvLUBack_64f; } /****************************************************************************************\ * Determinant of the matrix * \****************************************************************************************/ #define det2(m) (m(0,0)*m(1,1) - m(0,1)*m(1,0)) #define det3(m) (m(0,0)*(m(1,1)*m(2,2) - m(1,2)*m(2,1)) - \ m(0,1)*(m(1,0)*m(2,2) - m(1,2)*m(2,0)) + \ m(0,2)*(m(1,0)*m(2,1) - m(1,1)*m(2,0))) CV_IMPL double cvDet( const CvArr* arr ) { double result = 0; uchar* buffer = 0; int local_alloc = 0; CV_FUNCNAME( "cvDet" ); __BEGIN__; CvMat stub, *mat = (CvMat*)arr; int type; if( !CV_IS_MAT( mat )) { CV_CALL( mat = cvGetMat( mat, &stub )); } type = CV_MAT_TYPE( mat->type ); if( mat->width != mat->height ) CV_ERROR( CV_StsBadSize, "The matrix must be square" ); #define Mf( y, x ) ((float*)(m + y*step))[x] #define Md( y, x ) ((double*)(m + y*step))[x] if( mat->width == 2 ) { uchar* m = mat->data.ptr; int step = mat->step; if( type == CV_32FC1 ) { result = det2(Mf); } else if( type == CV_64FC1 ) { result = det2(Md); } else { CV_ERROR( CV_StsUnsupportedFormat, "" ); } } else if( mat->width == 3 ) { uchar* m = mat->data.ptr; int step = mat->step; if( type == CV_32FC1 ) { result = det3(Mf); } else if( type == CV_64FC1 ) { result = det3(Md); } else { CV_ERROR( CV_StsUnsupportedFormat, "" ); } } else if( mat->width == 1 ) { if( type == CV_32FC1 ) { result = mat->data.fl[0]; } else if( type == CV_64FC1 ) { result = mat->data.db[0]; } else { CV_ERROR( CV_StsUnsupportedFormat, "" ); } } else { CvLUDecompFunc decomp_func; CvSize size = cvGetMatSize( mat ); const int worktype = CV_64FC1; int buf_size = size.width*size.height*CV_ELEM_SIZE(worktype); CvMat tmat; if( !lu_inittab ) { icvInitLUTable( &lu_decomp_tab, &lu_back_tab ); lu_inittab = 1; } if( CV_MAT_CN( type ) != 1 || CV_MAT_DEPTH( type ) < CV_32F ) CV_ERROR( CV_StsUnsupportedFormat, "" ); if( size.width <= CV_MAX_LOCAL_MAT_SIZE ) { buffer = (uchar*)cvStackAlloc( buf_size ); local_alloc = 1; } else { CV_CALL( buffer = (uchar*)cvAlloc( buf_size )); } CV_CALL( cvInitMatHeader( &tmat, size.height, size.width, worktype, buffer )); if( type == worktype ) { CV_CALL( cvCopy( mat, &tmat )); } else CV_CALL( cvConvert( mat, &tmat )); decomp_func = (CvLUDecompFunc)(lu_decomp_tab.fn_2d[CV_MAT_DEPTH(worktype)-CV_32F]); assert( decomp_func ); IPPI_CALL( decomp_func( tmat.data.db, tmat.step, size, 0, 0, size, &result )); } #undef Mf #undef Md /*icvCheckVector_64f( &result, 1 );*/ __END__; if( buffer && !local_alloc ) cvFree( &buffer ); return result; } /****************************************************************************************\ * Inverse (or pseudo-inverse) of the matrix * \****************************************************************************************/ #define Sf( y, x ) ((float*)(srcdata + y*srcstep))[x] #define Sd( y, x ) ((double*)(srcdata + y*srcstep))[x] #define Df( y, x ) ((float*)(dstdata + y*dststep))[x] #define Dd( y, x ) ((double*)(dstdata + y*dststep))[x] CV_IMPL double cvInvert( const CvArr* srcarr, CvArr* dstarr, int method ) { CvMat* u = 0; CvMat* v = 0; CvMat* w = 0; uchar* buffer = 0; int local_alloc = 0; double result = 0; CV_FUNCNAME( "cvInvert" ); __BEGIN__; CvMat sstub, *src = (CvMat*)srcarr; CvMat dstub, *dst = (CvMat*)dstarr; int type; if( !CV_IS_MAT( src )) CV_CALL( src = cvGetMat( src, &sstub )); if( !CV_IS_MAT( dst )) CV_CALL( dst = cvGetMat( dst, &dstub )); type = CV_MAT_TYPE( src->type ); if( method == CV_SVD || method == CV_SVD_SYM ) { int n = MIN(src->rows,src->cols); if( method == CV_SVD_SYM && src->rows != src->cols ) CV_ERROR( CV_StsBadSize, "CV_SVD_SYM method is used for non-square matrix" ); CV_CALL( u = cvCreateMat( n, src->rows, src->type )); if( method != CV_SVD_SYM ) CV_CALL( v = cvCreateMat( n, src->cols, src->type )); CV_CALL( w = cvCreateMat( n, 1, src->type )); CV_CALL( cvSVD( src, w, u, v, CV_SVD_U_T + CV_SVD_V_T )); if( type == CV_32FC1 ) result = w->data.fl[0] >= FLT_EPSILON ? w->data.fl[w->rows-1]/w->data.fl[0] : 0; else result = w->data.db[0] >= FLT_EPSILON ? w->data.db[w->rows-1]/w->data.db[0] : 0; CV_CALL( cvSVBkSb( w, u, v ? v : u, 0, dst, CV_SVD_U_T + CV_SVD_V_T )); EXIT; } else if( method != CV_LU ) CV_ERROR( CV_StsBadArg, "Unknown inversion method" ); if( !CV_ARE_TYPES_EQ( src, dst )) CV_ERROR( CV_StsUnmatchedFormats, "" ); if( src->width != src->height ) CV_ERROR( CV_StsBadSize, "The matrix must be square" ); if( !CV_ARE_SIZES_EQ( src, dst )) CV_ERROR( CV_StsUnmatchedSizes, "" ); if( type != CV_32FC1 && type != CV_64FC1 ) CV_ERROR( CV_StsUnsupportedFormat, "" ); if( src->width <= 3 ) { uchar* srcdata = src->data.ptr; uchar* dstdata = dst->data.ptr; int srcstep = src->step; int dststep = dst->step; if( src->width == 2 ) { if( type == CV_32FC1 ) { double d = det2(Sf); if( d != 0. ) { double t0, t1; result = d; d = 1./d; t0 = Sf(0,0)*d; t1 = Sf(1,1)*d; Df(1,1) = (float)t0; Df(0,0) = (float)t1; t0 = -Sf(0,1)*d; t1 = -Sf(1,0)*d; Df(0,1) = (float)t0; Df(1,0) = (float)t1; } } else { double d = det2(Sd); if( d != 0. ) { double t0, t1; result = d; d = 1./d; t0 = Sd(0,0)*d; t1 = Sd(1,1)*d; Dd(1,1) = t0; Dd(0,0) = t1; t0 = -Sd(0,1)*d; t1 = -Sd(1,0)*d; Dd(0,1) = t0; Dd(1,0) = t1; } } } else if( src->width == 3 ) { if( type == CV_32FC1 ) { double d = det3(Sf); if( d != 0. ) { float t[9]; result = d; d = 1./d; t[0] = (float)((Sf(1,1) * Sf(2,2) - Sf(1,2) * Sf(2,1)) * d); t[1] = (float)((Sf(0,2) * Sf(2,1) - Sf(0,1) * Sf(2,2)) * d); t[2] = (float)((Sf(0,1) * Sf(1,2) - Sf(0,2) * Sf(1,1)) * d); t[3] = (float)((Sf(1,2) * Sf(2,0) - Sf(1,0) * Sf(2,2)) * d); t[4] = (float)((Sf(0,0) * Sf(2,2) - Sf(0,2) * Sf(2,0)) * d); t[5] = (float)((Sf(0,2) * Sf(1,0) - Sf(0,0) * Sf(1,2)) * d); t[6] = (float)((Sf(1,0) * Sf(2,1) - Sf(1,1) * Sf(2,0)) * d); t[7] = (float)((Sf(0,1) * Sf(2,0) - Sf(0,0) * Sf(2,1)) * d); t[8] = (float)((Sf(0,0) * Sf(1,1) - Sf(0,1) * Sf(1,0)) * d); Df(0,0) = t[0]; Df(0,1) = t[1]; Df(0,2) = t[2]; Df(1,0) = t[3]; Df(1,1) = t[4]; Df(1,2) = t[5]; Df(2,0) = t[6]; Df(2,1) = t[7]; Df(2,2) = t[8]; } } else { double d = det3(Sd); if( d != 0. ) { double t[9]; result = d; d = 1./d; t[0] = (Sd(1,1) * Sd(2,2) - Sd(1,2) * Sd(2,1)) * d; t[1] = (Sd(0,2) * Sd(2,1) - Sd(0,1) * Sd(2,2)) * d; t[2] = (Sd(0,1) * Sd(1,2) - Sd(0,2) * Sd(1,1)) * d; t[3] = (Sd(1,2) * Sd(2,0) - Sd(1,0) * Sd(2,2)) * d; t[4] = (Sd(0,0) * Sd(2,2) - Sd(0,2) * Sd(2,0)) * d; t[5] = (Sd(0,2) * Sd(1,0) - Sd(0,0) * Sd(1,2)) * d; t[6] = (Sd(1,0) * Sd(2,1) - Sd(1,1) * Sd(2,0)) * d; t[7] = (Sd(0,1) * Sd(2,0) - Sd(0,0) * Sd(2,1)) * d; t[8] = (Sd(0,0) * Sd(1,1) - Sd(0,1) * Sd(1,0)) * d; Dd(0,0) = t[0]; Dd(0,1) = t[1]; Dd(0,2) = t[2]; Dd(1,0) = t[3]; Dd(1,1) = t[4]; Dd(1,2) = t[5]; Dd(2,0) = t[6]; Dd(2,1) = t[7]; Dd(2,2) = t[8]; } } } else { assert( src->width == 1 ); if( type == CV_32FC1 ) { double d = Sf(0,0); if( d != 0. ) { result = d; Df(0,0) = (float)(1./d); } } else { double d = Sd(0,0); if( d != 0. ) { result = d; Dd(0,0) = 1./d; } } } } else { CvLUDecompFunc decomp_func; CvLUBackFunc back_func; CvSize size = cvGetMatSize( src ); const int worktype = CV_64FC1; int buf_size = size.width*size.height*CV_ELEM_SIZE(worktype); CvMat tmat; if( !lu_inittab ) { icvInitLUTable( &lu_decomp_tab, &lu_back_tab ); lu_inittab = 1; } if( size.width <= CV_MAX_LOCAL_MAT_SIZE ) { buffer = (uchar*)cvStackAlloc( buf_size ); local_alloc = 1; } else { CV_CALL( buffer = (uchar*)cvAlloc( buf_size )); } CV_CALL( cvInitMatHeader( &tmat, size.height, size.width, worktype, buffer )); if( type == worktype ) { CV_CALL( cvCopy( src, &tmat )); } else CV_CALL( cvConvert( src, &tmat )); CV_CALL( cvSetIdentity( dst )); decomp_func = (CvLUDecompFunc)(lu_decomp_tab.fn_2d[CV_MAT_DEPTH(type)-CV_32F]); back_func = (CvLUBackFunc)(lu_back_tab.fn_2d[CV_MAT_DEPTH(type)-CV_32F]); assert( decomp_func && back_func ); IPPI_CALL( decomp_func( tmat.data.db, tmat.step, size, dst->data.ptr, dst->step, size, &result )); if( result != 0 ) { IPPI_CALL( back_func( tmat.data.db, tmat.step, size, dst->data.ptr, dst->step, size )); } } if( !result ) CV_CALL( cvSetZero( dst )); __END__; if( buffer && !local_alloc ) cvFree( &buffer ); if( u || v || w ) { cvReleaseMat( &u ); cvReleaseMat( &v ); cvReleaseMat( &w ); } return result; } /****************************************************************************************\ * Linear system [least-squares] solution * \****************************************************************************************/ static void icvLSQ( const CvMat* A, const CvMat* B, CvMat* X ) { CvMat* AtA = 0; CvMat* AtB = 0; CvMat* W = 0; CvMat* V = 0; CV_FUNCNAME( "icvLSQ" ); __BEGIN__; if( !CV_IS_MAT(A) || !CV_IS_MAT(B) || !CV_IS_MAT(X) ) CV_ERROR( CV_StsBadArg, "Some of required arguments is not a valid matrix" ); AtA = cvCreateMat( A->cols, A->cols, A->type ); AtB = cvCreateMat( A->cols, 1, A->type ); W = cvCreateMat( A->cols, 1, A->type ); V = cvCreateMat( A->cols, A->cols, A->type ); cvMulTransposed( A, AtA, 1 ); cvGEMM( A, B, 1, 0, 0, AtB, CV_GEMM_A_T ); cvSVD( AtA, W, 0, V, CV_SVD_MODIFY_A + CV_SVD_V_T ); cvSVBkSb( W, V, V, AtB, X, CV_SVD_U_T + CV_SVD_V_T ); __END__; cvReleaseMat( &AtA ); cvReleaseMat( &AtB ); cvReleaseMat( &W ); cvReleaseMat( &V ); } CV_IMPL int cvSolve( const CvArr* A, const CvArr* b, CvArr* x, int method ) { CvMat* u = 0; CvMat* v = 0; CvMat* w = 0; uchar* buffer = 0; int local_alloc = 0; int result = 1; CV_FUNCNAME( "cvSolve" ); __BEGIN__; CvMat sstub, *src = (CvMat*)A; CvMat dstub, *dst = (CvMat*)x; CvMat bstub, *src2 = (CvMat*)b; int type; if( !CV_IS_MAT( src )) CV_CALL( src = cvGetMat( src, &sstub )); if( !CV_IS_MAT( src2 )) CV_CALL( src2 = cvGetMat( src2, &bstub )); if( !CV_IS_MAT( dst )) CV_CALL( dst = cvGetMat( dst, &dstub )); if( method & CV_LSQ ) { icvLSQ( src, src2, dst ); EXIT; } if( method == CV_SVD || method == CV_SVD_SYM ) { int n = MIN(src->rows,src->cols); if( method == CV_SVD_SYM && src->rows != src->cols ) CV_ERROR( CV_StsBadSize, "CV_SVD_SYM method is used for non-square matrix" ); CV_CALL( u = cvCreateMat( n, src->rows, src->type )); if( method != CV_SVD_SYM ) CV_CALL( v = cvCreateMat( n, src->cols, src->type )); CV_CALL( w = cvCreateMat( n, 1, src->type )); CV_CALL( cvSVD( src, w, u, v, CV_SVD_U_T + CV_SVD_V_T )); CV_CALL( cvSVBkSb( w, u, v ? v : u, src2, dst, CV_SVD_U_T + CV_SVD_V_T )); EXIT; } else if( method != CV_LU ) CV_ERROR( CV_StsBadArg, "Unknown inversion method" ); type = CV_MAT_TYPE( src->type ); if( !CV_ARE_TYPES_EQ( src, dst ) || !CV_ARE_TYPES_EQ( src, src2 )) CV_ERROR( CV_StsUnmatchedFormats, "" ); if( src->width != src->height ) CV_ERROR( CV_StsBadSize, "The matrix must be square" ); if( !CV_ARE_SIZES_EQ( src2, dst ) || src->width != src2->height ) CV_ERROR( CV_StsUnmatchedSizes, "" ); if( type != CV_32FC1 && type != CV_64FC1 ) CV_ERROR( CV_StsUnsupportedFormat, "" ); // check case of a single equation and small matrix if( src->width <= 3 && src2->width == 1 ) { #define bf(y) ((float*)(bdata + y*src2step))[0] #define bd(y) ((double*)(bdata + y*src2step))[0] uchar* srcdata = src->data.ptr; uchar* bdata = src2->data.ptr; uchar* dstdata = dst->data.ptr; int srcstep = src->step; int src2step = src2->step; int dststep = dst->step; if( src->width == 2 ) { if( type == CV_32FC1 ) { double d = det2(Sf); if( d != 0. ) { float t; d = 1./d; t = (float)((bf(0)*Sf(1,1) - bf(1)*Sf(0,1))*d); Df(1,0) = (float)((bf(1)*Sf(0,0) - bf(0)*Sf(1,0))*d); Df(0,0) = t; } else result = 0; } else { double d = det2(Sd); if( d != 0. ) { double t; d = 1./d; t = (bd(0)*Sd(1,1) - bd(1)*Sd(0,1))*d; Dd(1,0) = (bd(1)*Sd(0,0) - bd(0)*Sd(1,0))*d; Dd(0,0) = t; } else result = 0; } } else if( src->width == 3 ) { if( type == CV_32FC1 ) { double d = det3(Sf); if( d != 0. ) { float t[3]; d = 1./d; t[0] = (float)(d* (bf(0)*(Sf(1,1)*Sf(2,2) - Sf(1,2)*Sf(2,1)) - Sf(0,1)*(bf(1)*Sf(2,2) - Sf(1,2)*bf(2)) + Sf(0,2)*(bf(1)*Sf(2,1) - Sf(1,1)*bf(2)))); t[1] = (float)(d* (Sf(0,0)*(bf(1)*Sf(2,2) - Sf(1,2)*bf(2)) - bf(0)*(Sf(1,0)*Sf(2,2) - Sf(1,2)*Sf(2,0)) + Sf(0,2)*(Sf(1,0)*bf(2) - bf(1)*Sf(2,0)))); t[2] = (float)(d* (Sf(0,0)*(Sf(1,1)*bf(2) - bf(1)*Sf(2,1)) - Sf(0,1)*(Sf(1,0)*bf(2) - bf(1)*Sf(2,0)) + bf(0)*(Sf(1,0)*Sf(2,1) - Sf(1,1)*Sf(2,0)))); Df(0,0) = t[0]; Df(1,0) = t[1]; Df(2,0) = t[2]; } else result = 0; } else { double d = det3(Sd); if( d != 0. ) { double t[9]; d = 1./d; t[0] = ((Sd(1,1) * Sd(2,2) - Sd(1,2) * Sd(2,1))*bd(0) + (Sd(0,2) * Sd(2,1) - Sd(0,1) * Sd(2,2))*bd(1) + (Sd(0,1) * Sd(1,2) - Sd(0,2) * Sd(1,1))*bd(2))*d; t[1] = ((Sd(1,2) * Sd(2,0) - Sd(1,0) * Sd(2,2))*bd(0) + (Sd(0,0) * Sd(2,2) - Sd(0,2) * Sd(2,0))*bd(1) + (Sd(0,2) * Sd(1,0) - Sd(0,0) * Sd(1,2))*bd(2))*d; t[2] = ((Sd(1,0) * Sd(2,1) - Sd(1,1) * Sd(2,0))*bd(0) + (Sd(0,1) * Sd(2,0) - Sd(0,0) * Sd(2,1))*bd(1) + (Sd(0,0) * Sd(1,1) - Sd(0,1) * Sd(1,0))*bd(2))*d; Dd(0,0) = t[0]; Dd(1,0) = t[1]; Dd(2,0) = t[2]; } else result = 0; } } else { assert( src->width == 1 ); if( type == CV_32FC1 ) { double d = Sf(0,0); if( d != 0. ) Df(0,0) = (float)(bf(0)/d); else result = 0; } else { double d = Sd(0,0); if( d != 0. ) Dd(0,0) = (bd(0)/d); else result = 0; } } } else { CvLUDecompFunc decomp_func; CvLUBackFunc back_func; CvSize size = cvGetMatSize( src ); CvSize dstsize = cvGetMatSize( dst ); int worktype = CV_64FC1; int buf_size = size.width*size.height*CV_ELEM_SIZE(worktype); double d = 0; CvMat tmat; if( !lu_inittab ) { icvInitLUTable( &lu_decomp_tab, &lu_back_tab ); lu_inittab = 1; } if( size.width <= CV_MAX_LOCAL_MAT_SIZE ) { buffer = (uchar*)cvStackAlloc( buf_size ); local_alloc = 1; } else { CV_CALL( buffer = (uchar*)cvAlloc( buf_size )); } CV_CALL( cvInitMatHeader( &tmat, size.height, size.width, worktype, buffer )); if( type == worktype ) { CV_CALL( cvCopy( src, &tmat )); } else CV_CALL( cvConvert( src, &tmat )); if( src2->data.ptr != dst->data.ptr ) { CV_CALL( cvCopy( src2, dst )); } decomp_func = (CvLUDecompFunc)(lu_decomp_tab.fn_2d[CV_MAT_DEPTH(type)-CV_32F]); back_func = (CvLUBackFunc)(lu_back_tab.fn_2d[CV_MAT_DEPTH(type)-CV_32F]); assert( decomp_func && back_func ); IPPI_CALL( decomp_func( tmat.data.db, tmat.step, size, dst->data.ptr, dst->step, dstsize, &d )); if( d != 0 ) { IPPI_CALL( back_func( tmat.data.db, tmat.step, size, dst->data.ptr, dst->step, dstsize )); } else result = 0; } if( !result ) CV_CALL( cvSetZero( dst )); __END__; if( buffer && !local_alloc ) cvFree( &buffer ); if( u || v || w ) { cvReleaseMat( &u ); cvReleaseMat( &v ); cvReleaseMat( &w ); } return result; } /****************************************************************************************\ * 3D vector cross-product * \****************************************************************************************/ CV_IMPL void cvCrossProduct( const CvArr* srcAarr, const CvArr* srcBarr, CvArr* dstarr ) { CV_FUNCNAME( "cvCrossProduct" ); __BEGIN__; CvMat stubA, *srcA = (CvMat*)srcAarr; CvMat stubB, *srcB = (CvMat*)srcBarr; CvMat dstub, *dst = (CvMat*)dstarr; int type; if( !CV_IS_MAT(srcA)) CV_CALL( srcA = cvGetMat( srcA, &stubA )); type = CV_MAT_TYPE( srcA->type ); if( srcA->width*srcA->height*CV_MAT_CN(type) != 3 ) CV_ERROR( CV_StsBadArg, "All the input arrays must be continuous 3-vectors" ); if( !srcB || !dst ) CV_ERROR( CV_StsNullPtr, "" ); if( (srcA->type & ~CV_MAT_CONT_FLAG) == (srcB->type & ~CV_MAT_CONT_FLAG) && (srcA->type & ~CV_MAT_CONT_FLAG) == (dst->type & ~CV_MAT_CONT_FLAG) ) { if( !srcB->data.ptr || !dst->data.ptr ) CV_ERROR( CV_StsNullPtr, "" ); } else { if( !CV_IS_MAT(srcB)) CV_CALL( srcB = cvGetMat( srcB, &stubB )); if( !CV_IS_MAT(dst)) CV_CALL( dst = cvGetMat( dst, &dstub )); if( !CV_ARE_TYPES_EQ( srcA, srcB ) || !CV_ARE_TYPES_EQ( srcB, dst )) CV_ERROR( CV_StsUnmatchedFormats, "" ); } if( !CV_ARE_SIZES_EQ( srcA, srcB ) || !CV_ARE_SIZES_EQ( srcB, dst )) CV_ERROR( CV_StsUnmatchedSizes, "" ); if( CV_MAT_DEPTH(type) == CV_32F ) { float* dstdata = (float*)(dst->data.ptr); const float* src1data = (float*)(srcA->data.ptr); const float* src2data = (float*)(srcB->data.ptr); if( CV_IS_MAT_CONT(srcA->type & srcB->type & dst->type) ) { dstdata[2] = src1data[0] * src2data[1] - src1data[1] * src2data[0]; dstdata[0] = src1data[1] * src2data[2] - src1data[2] * src2data[1]; dstdata[1] = src1data[2] * src2data[0] - src1data[0] * src2data[2]; } else { int step1 = srcA->step ? srcA->step/sizeof(src1data[0]) : 1; int step2 = srcB->step ? srcB->step/sizeof(src1data[0]) : 1; int step = dst->step ? dst->step/sizeof(src1data[0]) : 1; dstdata[2*step] = src1data[0] * src2data[step2] - src1data[step1] * src2data[0]; dstdata[0] = src1data[step1] * src2data[step2*2] - src1data[step1*2] * src2data[step2]; dstdata[step] = src1data[step1*2] * src2data[0] - src1data[0] * src2data[step2*2]; } } else if( CV_MAT_DEPTH(type) == CV_64F ) { double* dstdata = (double*)(dst->data.ptr); const double* src1data = (double*)(srcA->data.ptr); const double* src2data = (double*)(srcB->data.ptr); if( CV_IS_MAT_CONT(srcA->type & srcB->type & dst->type) ) { dstdata[2] = src1data[0] * src2data[1] - src1data[1] * src2data[0]; dstdata[0] = src1data[1] * src2data[2] - src1data[2] * src2data[1]; dstdata[1] = src1data[2] * src2data[0] - src1data[0] * src2data[2]; } else { int step1 = srcA->step ? srcA->step/sizeof(src1data[0]) : 1; int step2 = srcB->step ? srcB->step/sizeof(src1data[0]) : 1; int step = dst->step ? dst->step/sizeof(src1data[0]) : 1; dstdata[2*step] = src1data[0] * src2data[step2] - src1data[step1] * src2data[0]; dstdata[0] = src1data[step1] * src2data[step2*2] - src1data[step1*2] * src2data[step2]; dstdata[step] = src1data[step1*2] * src2data[0] - src1data[0] * src2data[step2*2]; } } else { CV_ERROR( CV_StsUnsupportedFormat, "" ); } __END__; } CV_IMPL void cvCalcPCA( const CvArr* data_arr, CvArr* avg_arr, CvArr* eigenvals, CvArr* eigenvects, int flags ) { CvMat* tmp_avg = 0; CvMat* tmp_avg_r = 0; CvMat* tmp_cov = 0; CvMat* tmp_evals = 0; CvMat* tmp_evects = 0; CvMat* tmp_evects2 = 0; CvMat* tmp_data = 0; CV_FUNCNAME( "cvCalcPCA" ); __BEGIN__; CvMat stub, *data = (CvMat*)data_arr; CvMat astub, *avg = (CvMat*)avg_arr; CvMat evalstub, *evals = (CvMat*)eigenvals; CvMat evectstub, *evects = (CvMat*)eigenvects; int covar_flags = CV_COVAR_SCALE; int i, len, in_count, count, out_count; if( !CV_IS_MAT(data) ) CV_CALL( data = cvGetMat( data, &stub )); if( !CV_IS_MAT(avg) ) CV_CALL( avg = cvGetMat( avg, &astub )); if( !CV_IS_MAT(evals) ) CV_CALL( evals = cvGetMat( evals, &evalstub )); if( !CV_IS_MAT(evects) ) CV_CALL( evects = cvGetMat( evects, &evectstub )); if( CV_MAT_CN(data->type) != 1 || CV_MAT_CN(avg->type) != 1 || CV_MAT_CN(evals->type) != 1 || CV_MAT_CN(evects->type) != 1 ) CV_ERROR( CV_StsUnsupportedFormat, "All the input and output arrays must be 1-channel" ); if( CV_MAT_DEPTH(avg->type) < CV_32F || !CV_ARE_DEPTHS_EQ(avg, evals) || !CV_ARE_DEPTHS_EQ(avg, evects) ) CV_ERROR( CV_StsUnsupportedFormat, "All the output arrays must have the same type, 32fC1 or 64fC1" ); if( flags & CV_PCA_DATA_AS_COL ) { len = data->rows; in_count = data->cols; covar_flags |= CV_COVAR_COLS; if( avg->cols != 1 || avg->rows != len ) CV_ERROR( CV_StsBadSize, "The mean (average) vector should be data->rows x 1 when CV_PCA_DATA_AS_COL is used" ); CV_CALL( tmp_avg = cvCreateMat( len, 1, CV_64F )); } else { len = data->cols; in_count = data->rows; covar_flags |= CV_COVAR_ROWS; if( avg->rows != 1 || avg->cols != len ) CV_ERROR( CV_StsBadSize, "The mean (average) vector should be 1 x data->cols when CV_PCA_DATA_AS_ROW is used" ); CV_CALL( tmp_avg = cvCreateMat( 1, len, CV_64F )); } count = MIN(len, in_count); out_count = evals->cols + evals->rows - 1; if( (evals->cols != 1 && evals->rows != 1) || out_count > count ) CV_ERROR( CV_StsBadSize, "The array of eigenvalues must be 1d vector containing " "no more than min(data->rows,data->cols) elements" ); if( evects->cols != len || evects->rows != out_count ) CV_ERROR( CV_StsBadSize, "The matrix of eigenvalues must have the same number of columns as the input vector length " "and the same number of rows as the number of eigenvalues" ); // "scrambled" way to compute PCA (when cols(A)>rows(A)): // B = A'A; B*x=b*x; C = AA'; C*y=c*y -> AA'*y=c*y -> A'A*(A'*y)=c*(A'*y) -> c = b, x=A'*y if( len <= in_count ) covar_flags |= CV_COVAR_NORMAL; if( flags & CV_PCA_USE_AVG ){ covar_flags |= CV_COVAR_USE_AVG; CV_CALL( cvConvert( avg, tmp_avg ) ); } CV_CALL( tmp_cov = cvCreateMat( count, count, CV_64F )); CV_CALL( tmp_evals = cvCreateMat( 1, count, CV_64F )); CV_CALL( tmp_evects = cvCreateMat( count, count, CV_64F )); CV_CALL( cvCalcCovarMatrix( &data_arr, 0, tmp_cov, tmp_avg, covar_flags )); CV_CALL( cvSVD( tmp_cov, tmp_evals, tmp_evects, 0, CV_SVD_MODIFY_A + CV_SVD_U_T )); tmp_evects->rows = out_count; tmp_evals->cols = out_count; cvZero( evects ); cvZero( evals ); if( covar_flags & CV_COVAR_NORMAL ) { CV_CALL( cvConvert( tmp_evects, evects )); } else { // CV_PCA_DATA_AS_ROW: cols(A)>rows(A). x=A'*y -> x'=y'*A // CV_PCA_DATA_AS_COL: rows(A)>cols(A). x=A''*y -> x'=y'*A' int block_count = 0; CV_CALL( tmp_data = cvCreateMat( count, count, CV_64F )); CV_CALL( tmp_avg_r = cvCreateMat( count, count, CV_64F )); CV_CALL( tmp_evects2 = cvCreateMat( out_count, count, CV_64F )); for( i = 0; i < len; i += block_count ) { CvMat data_part, tdata_part, part, dst_part, avg_part, tmp_avg_part; int gemm_flags; block_count = MIN( count, len - i ); if( flags & CV_PCA_DATA_AS_COL ) { cvGetRows( data, &data_part, i, i + block_count ); cvGetRows( tmp_data, &tdata_part, 0, block_count ); cvGetRows( tmp_avg, &avg_part, i, i + block_count ); cvGetRows( tmp_avg_r, &tmp_avg_part, 0, block_count ); gemm_flags = CV_GEMM_B_T; } else { cvGetCols( data, &data_part, i, i + block_count ); cvGetCols( tmp_data, &tdata_part, 0, block_count ); cvGetCols( tmp_avg, &avg_part, i, i + block_count ); cvGetCols( tmp_avg_r, &tmp_avg_part, 0, block_count ); gemm_flags = 0; } cvGetCols( tmp_evects2, &part, 0, block_count ); cvGetCols( evects, &dst_part, i, i + block_count ); cvConvert( &data_part, &tdata_part ); cvRepeat( &avg_part, &tmp_avg_part ); cvSub( &tdata_part, &tmp_avg_part, &tdata_part ); cvGEMM( tmp_evects, &tdata_part, 1, 0, 0, &part, gemm_flags ); cvConvert( &part, &dst_part ); } // normalize eigenvectors for( i = 0; i < out_count; i++ ) { CvMat ei; cvGetRow( evects, &ei, i ); cvNormalize( &ei, &ei ); } } if( tmp_evals->rows != evals->rows ) cvReshape( tmp_evals, tmp_evals, 1, evals->rows ); cvConvert( tmp_evals, evals ); cvConvert( tmp_avg, avg ); __END__; cvReleaseMat( &tmp_avg ); cvReleaseMat( &tmp_avg_r ); cvReleaseMat( &tmp_cov ); cvReleaseMat( &tmp_evals ); cvReleaseMat( &tmp_evects ); cvReleaseMat( &tmp_evects2 ); cvReleaseMat( &tmp_data ); } CV_IMPL void cvProjectPCA( const CvArr* data_arr, const CvArr* avg_arr, const CvArr* eigenvects, CvArr* result_arr ) { uchar* buffer = 0; int local_alloc = 0; CV_FUNCNAME( "cvProjectPCA" ); __BEGIN__; CvMat stub, *data = (CvMat*)data_arr; CvMat astub, *avg = (CvMat*)avg_arr; CvMat evectstub, *evects = (CvMat*)eigenvects; CvMat rstub, *result = (CvMat*)result_arr; CvMat avg_repeated; int i, len, in_count; int gemm_flags, as_cols, convert_data; int block_count0, block_count, buf_size, elem_size; uchar* tmp_data_ptr; if( !CV_IS_MAT(data) ) CV_CALL( data = cvGetMat( data, &stub )); if( !CV_IS_MAT(avg) ) CV_CALL( avg = cvGetMat( avg, &astub )); if( !CV_IS_MAT(evects) ) CV_CALL( evects = cvGetMat( evects, &evectstub )); if( !CV_IS_MAT(result) ) CV_CALL( result = cvGetMat( result, &rstub )); if( CV_MAT_CN(data->type) != 1 || CV_MAT_CN(avg->type) != 1 ) CV_ERROR( CV_StsUnsupportedFormat, "All the input and output arrays must be 1-channel" ); if( (CV_MAT_TYPE(avg->type) != CV_32FC1 && CV_MAT_TYPE(avg->type) != CV_64FC1) || !CV_ARE_TYPES_EQ(avg, evects) || !CV_ARE_TYPES_EQ(avg, result) ) CV_ERROR( CV_StsUnsupportedFormat, "All the input and output arrays (except for data) must have the same type, 32fC1 or 64fC1" ); if( (avg->cols != 1 || avg->rows != data->rows) && (avg->rows != 1 || avg->cols != data->cols) ) CV_ERROR( CV_StsBadSize, "The mean (average) vector should be either 1 x data->cols or data->rows x 1" ); if( avg->cols == 1 ) { len = data->rows; in_count = data->cols; gemm_flags = CV_GEMM_A_T + CV_GEMM_B_T; as_cols = 1; } else { len = data->cols; in_count = data->rows; gemm_flags = CV_GEMM_B_T; as_cols = 0; } if( evects->cols != len ) CV_ERROR( CV_StsUnmatchedSizes, "Eigenvectors must be stored as rows and be of the same size as input vectors" ); if( result->cols > evects->rows ) CV_ERROR( CV_StsOutOfRange, "The output matrix of coefficients must have the number of columns " "less than or equal to the number of eigenvectors (number of rows in eigenvectors matrix)" ); evects = cvGetRows( evects, &evectstub, 0, result->cols ); block_count0 = (1 << 16)/len; block_count0 = MAX( block_count0, 4 ); block_count0 = MIN( block_count0, in_count ); elem_size = CV_ELEM_SIZE(avg->type); convert_data = CV_MAT_DEPTH(data->type) < CV_MAT_DEPTH(avg->type); buf_size = block_count0*len*((block_count0 > 1) + 1)*elem_size; if( buf_size < CV_MAX_LOCAL_SIZE ) { buffer = (uchar*)cvStackAlloc( buf_size ); local_alloc = 1; } else CV_CALL( buffer = (uchar*)cvAlloc( buf_size )); tmp_data_ptr = buffer; if( block_count0 > 1 ) { avg_repeated = cvMat( as_cols ? len : block_count0, as_cols ? block_count0 : len, avg->type, buffer ); cvRepeat( avg, &avg_repeated ); tmp_data_ptr += block_count0*len*elem_size; } else avg_repeated = *avg; for( i = 0; i < in_count; i += block_count ) { CvMat data_part, norm_data, avg_part, *src = &data_part, out_part; block_count = MIN( block_count0, in_count - i ); if( as_cols ) { cvGetCols( data, &data_part, i, i + block_count ); cvGetCols( &avg_repeated, &avg_part, 0, block_count ); norm_data = cvMat( len, block_count, avg->type, tmp_data_ptr ); } else { cvGetRows( data, &data_part, i, i + block_count ); cvGetRows( &avg_repeated, &avg_part, 0, block_count ); norm_data = cvMat( block_count, len, avg->type, tmp_data_ptr ); } if( convert_data ) { cvConvert( src, &norm_data ); src = &norm_data; } cvSub( src, &avg_part, &norm_data ); cvGetRows( result, &out_part, i, i + block_count ); cvGEMM( &norm_data, evects, 1, 0, 0, &out_part, gemm_flags ); } __END__; if( !local_alloc ) cvFree( &buffer ); } CV_IMPL void cvBackProjectPCA( const CvArr* proj_arr, const CvArr* avg_arr, const CvArr* eigenvects, CvArr* result_arr ) { uchar* buffer = 0; int local_alloc = 0; CV_FUNCNAME( "cvProjectPCA" ); __BEGIN__; CvMat pstub, *data = (CvMat*)proj_arr; CvMat astub, *avg = (CvMat*)avg_arr; CvMat evectstub, *evects = (CvMat*)eigenvects; CvMat rstub, *result = (CvMat*)result_arr; CvMat avg_repeated; int i, len, in_count, as_cols; int block_count0, block_count, buf_size, elem_size; if( !CV_IS_MAT(data) ) CV_CALL( data = cvGetMat( data, &pstub )); if( !CV_IS_MAT(avg) ) CV_CALL( avg = cvGetMat( avg, &astub )); if( !CV_IS_MAT(evects) ) CV_CALL( evects = cvGetMat( evects, &evectstub )); if( !CV_IS_MAT(result) ) CV_CALL( result = cvGetMat( result, &rstub )); if( (CV_MAT_TYPE(avg->type) != CV_32FC1 && CV_MAT_TYPE(avg->type) != CV_64FC1) || !CV_ARE_TYPES_EQ(avg, data) || !CV_ARE_TYPES_EQ(avg, evects) || !CV_ARE_TYPES_EQ(avg, result) ) CV_ERROR( CV_StsUnsupportedFormat, "All the input and output arrays must have the same type, 32fC1 or 64fC1" ); if( (avg->cols != 1 || avg->rows != result->rows) && (avg->rows != 1 || avg->cols != result->cols) ) CV_ERROR( CV_StsBadSize, "The mean (average) vector should be either 1 x result->cols or result->rows x 1" ); if( avg->cols == 1 ) { len = result->rows; in_count = result->cols; as_cols = 1; } else { len = result->cols; in_count = result->rows; as_cols = 0; } if( evects->cols != len ) CV_ERROR( CV_StsUnmatchedSizes, "Eigenvectors must be stored as rows and be of the same size as the output vectors" ); if( data->cols > evects->rows ) CV_ERROR( CV_StsOutOfRange, "The input matrix of coefficients must have the number of columns " "less than or equal to the number of eigenvectors (number of rows in eigenvectors matrix)" ); evects = cvGetRows( evects, &evectstub, 0, data->cols ); block_count0 = (1 << 16)/len; block_count0 = MAX( block_count0, 4 ); block_count0 = MIN( block_count0, in_count ); elem_size = CV_ELEM_SIZE(avg->type); buf_size = block_count0*len*(block_count0 > 1)*elem_size; if( buf_size < CV_MAX_LOCAL_SIZE ) { buffer = (uchar*)cvStackAlloc( MAX(buf_size,16) ); local_alloc = 1; } else CV_CALL( buffer = (uchar*)cvAlloc( buf_size )); if( block_count0 > 1 ) { avg_repeated = cvMat( as_cols ? len : block_count0, as_cols ? block_count0 : len, avg->type, buffer ); cvRepeat( avg, &avg_repeated ); } else avg_repeated = *avg; for( i = 0; i < in_count; i += block_count ) { CvMat data_part, avg_part, out_part; block_count = MIN( block_count0, in_count - i ); cvGetRows( data, &data_part, i, i + block_count ); if( as_cols ) { cvGetCols( result, &out_part, i, i + block_count ); cvGetCols( &avg_repeated, &avg_part, 0, block_count ); cvGEMM( evects, &data_part, 1, &avg_part, 1, &out_part, CV_GEMM_A_T + CV_GEMM_B_T ); } else { cvGetRows( result, &out_part, i, i + block_count ); cvGetRows( &avg_repeated, &avg_part, 0, block_count ); cvGEMM( &data_part, evects, 1, &avg_part, 1, &out_part, 0 ); } } __END__; if( !local_alloc ) cvFree( &buffer ); } /* End of file. */