/*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*/
/* ////////////////////////////////////////////////////////////////////
//
// Filling CvMat/IplImage instances with random numbers
//
// */
#include "_cxcore.h"
///////////////////////////// Functions Declaration //////////////////////////////////////
/*
Multiply-with-carry generator is used here:
temp = ( A*X(n) + carry )
X(n+1) = temp mod (2^32)
carry = temp / (2^32)
*/
#define ICV_RNG_NEXT(x) ((uint64)(unsigned)(x)*1554115554 + ((x) >> 32))
#define ICV_CVT_FLT(x) (((unsigned)(x) >> 9)|CV_1F)
#define ICV_1D CV_BIG_INT(0x3FF0000000000000)
#define ICV_CVT_DBL(x) (((uint64)(unsigned)(x) << 20)|((x) >> 44)|ICV_1D)
/***************************************************************************************\
* Pseudo-Random Number Generators (PRNGs) *
\***************************************************************************************/
#define ICV_IMPL_RAND_BITS( flavor, arrtype, cast_macro ) \
static CvStatus CV_STDCALL \
icvRandBits_##flavor##_C1R( arrtype* arr, int step, CvSize size, \
uint64* state, const int* param ) \
{ \
uint64 temp = *state; \
int small_flag = (param[12]|param[13]|param[14]|param[15]) <= 255; \
step /= sizeof(arr[0]); \
\
for( ; size.height--; arr += step ) \
{ \
int i, k = 3; \
const int* p = param; \
\
if( !small_flag ) \
{ \
for( i = 0; i <= size.width - 4; i += 4 ) \
{ \
unsigned t0, t1; \
\
temp = ICV_RNG_NEXT(temp); \
t0 = ((unsigned)temp & p[i + 12]) + p[i]; \
temp = ICV_RNG_NEXT(temp); \
t1 = ((unsigned)temp & p[i + 13]) + p[i+1]; \
arr[i] = cast_macro((int)t0); \
arr[i+1] = cast_macro((int)t1); \
\
temp = ICV_RNG_NEXT(temp); \
t0 = ((unsigned)temp & p[i + 14]) + p[i+2]; \
temp = ICV_RNG_NEXT(temp); \
t1 = ((unsigned)temp & p[i + 15]) + p[i+3]; \
arr[i+2] = cast_macro((int)t0); \
arr[i+3] = cast_macro((int)t1); \
\
if( !--k ) \
{ \
k = 3; \
p -= 12; \
} \
} \
} \
else \
{ \
for( i = 0; i <= size.width - 4; i += 4 ) \
{ \
unsigned t0, t1, t; \
\
temp = ICV_RNG_NEXT(temp); \
t = (unsigned)temp; \
t0 = (t & p[i + 12]) + p[i]; \
t1 = ((t >> 8) & p[i + 13]) + p[i+1]; \
arr[i] = cast_macro((int)t0); \
arr[i+1] = cast_macro((int)t1); \
\
t0 = ((t >> 16) & p[i + 14]) + p[i + 2]; \
t1 = ((t >> 24) & p[i + 15]) + p[i + 3]; \
arr[i+2] = cast_macro((int)t0); \
arr[i+3] = cast_macro((int)t1); \
\
if( !--k ) \
{ \
k = 3; \
p -= 12; \
} \
} \
} \
\
for( ; i < size.width; i++ ) \
{ \
unsigned t0; \
temp = ICV_RNG_NEXT(temp); \
\
t0 = ((unsigned)temp & p[i + 12]) + p[i]; \
arr[i] = cast_macro((int)t0); \
} \
} \
\
*state = temp; \
return CV_OK; \
}
#define ICV_IMPL_RAND( flavor, arrtype, worktype, cast_macro1, cast_macro2 )\
static CvStatus CV_STDCALL \
icvRand_##flavor##_C1R( arrtype* arr, int step, CvSize size, \
uint64* state, const double* param ) \
{ \
uint64 temp = *state; \
step /= sizeof(arr[0]); \
\
for( ; size.height--; arr += step ) \
{ \
int i, k = 3; \
const double* p = param; \
\
for( i = 0; i <= size.width - 4; i += 4 ) \
{ \
worktype f0, f1; \
Cv32suf t0, t1; \
\
temp = ICV_RNG_NEXT(temp); \
t0.u = ICV_CVT_FLT(temp); \
temp = ICV_RNG_NEXT(temp); \
t1.u = ICV_CVT_FLT(temp); \
f0 = cast_macro1( t0.f * p[i + 12] + p[i] ); \
f1 = cast_macro1( t1.f * p[i + 13] + p[i + 1] ); \
arr[i] = cast_macro2(f0); \
arr[i+1] = cast_macro2(f1); \
\
temp = ICV_RNG_NEXT(temp); \
t0.u = ICV_CVT_FLT(temp); \
temp = ICV_RNG_NEXT(temp); \
t1.u = ICV_CVT_FLT(temp); \
f0 = cast_macro1( t0.f * p[i + 14] + p[i + 2] ); \
f1 = cast_macro1( t1.f * p[i + 15] + p[i + 3] ); \
arr[i+2] = cast_macro2(f0); \
arr[i+3] = cast_macro2(f1); \
\
if( !--k ) \
{ \
k = 3; \
p -= 12; \
} \
} \
\
for( ; i < size.width; i++ ) \
{ \
worktype f0; \
Cv32suf t0; \
\
temp = ICV_RNG_NEXT(temp); \
t0.u = ICV_CVT_FLT(temp); \
f0 = cast_macro1( t0.f * p[i + 12] + p[i] ); \
arr[i] = cast_macro2(f0); \
} \
} \
\
*state = temp; \
return CV_OK; \
}
static CvStatus CV_STDCALL
icvRand_64f_C1R( double* arr, int step, CvSize size,
uint64* state, const double* param )
{
uint64 temp = *state;
step /= sizeof(arr[0]);
for( ; size.height--; arr += step )
{
int i, k = 3;
const double* p = param;
for( i = 0; i <= size.width - 4; i += 4 )
{
double f0, f1;
Cv64suf t0, t1;
temp = ICV_RNG_NEXT(temp);
t0.u = ICV_CVT_DBL(temp);
temp = ICV_RNG_NEXT(temp);
t1.u = ICV_CVT_DBL(temp);
f0 = t0.f * p[i + 12] + p[i];
f1 = t1.f * p[i + 13] + p[i + 1];
arr[i] = f0;
arr[i+1] = f1;
temp = ICV_RNG_NEXT(temp);
t0.u = ICV_CVT_DBL(temp);
temp = ICV_RNG_NEXT(temp);
t1.u = ICV_CVT_DBL(temp);
f0 = t0.f * p[i + 14] + p[i + 2];
f1 = t1.f * p[i + 15] + p[i + 3];
arr[i+2] = f0;
arr[i+3] = f1;
if( !--k )
{
k = 3;
p -= 12;
}
}
for( ; i < size.width; i++ )
{
double f0;
Cv64suf t0;
temp = ICV_RNG_NEXT(temp);
t0.u = ICV_CVT_DBL(temp);
f0 = t0.f * p[i + 12] + p[i];
arr[i] = f0;
}
}
*state = temp;
return CV_OK;
}
/***************************************************************************************\
The code below implements algorithm from the paper:
G. Marsaglia and W.W. Tsang,
The Monty Python method for generating random variables,
ACM Transactions on Mathematical Software, Vol. 24, No. 3,
Pages 341-350, September, 1998.
\***************************************************************************************/
static CvStatus CV_STDCALL
icvRandn_0_1_32f_C1R( float* arr, int len, uint64* state )
{
uint64 temp = *state;
int i;
temp = ICV_RNG_NEXT(temp);
for( i = 0; i < len; i++ )
{
double x, y, v, ax, bx;
for(;;)
{
x = ((int)temp)*1.167239e-9;
temp = ICV_RNG_NEXT(temp);
ax = fabs(x);
v = 2.8658 - ax*(2.0213 - 0.3605*ax);
y = ((unsigned)temp)*2.328306e-10;
temp = ICV_RNG_NEXT(temp);
if( y < v || ax < 1.17741 )
break;
bx = x;
x = bx > 0 ? 0.8857913*(2.506628 - ax) : -0.8857913*(2.506628 - ax);
if( y > v + 0.0506 )
break;
if( log(y) < .6931472 - .5*bx*bx )
{
x = bx;
break;
}
if( log(1.8857913 - y) < .5718733-.5*x*x )
break;
do
{
v = ((int)temp)*4.656613e-10;
x = -log(fabs(v))*.3989423;
temp = ICV_RNG_NEXT(temp);
y = -log(((unsigned)temp)*2.328306e-10);
temp = ICV_RNG_NEXT(temp);
}
while( y+y < x*x );
x = v > 0 ? 2.506628 + x : -2.506628 - x;
break;
}
arr[i] = (float)x;
}
*state = temp;
return CV_OK;
}
#define RAND_BUF_SIZE 96
#define ICV_IMPL_RANDN( flavor, arrtype, worktype, cast_macro1, cast_macro2 ) \
static CvStatus CV_STDCALL \
icvRandn_##flavor##_C1R( arrtype* arr, int step, CvSize size, \
uint64* state, const double* param ) \
{ \
float buffer[RAND_BUF_SIZE]; \
step /= sizeof(arr[0]); \
\
for( ; size.height--; arr += step ) \
{ \
int i, j, len = RAND_BUF_SIZE; \
\
for( i = 0; i < size.width; i += RAND_BUF_SIZE ) \
{ \
int k = 3; \
const double* p = param; \
\
if( i + len > size.width ) \
len = size.width - i; \
\
icvRandn_0_1_32f_C1R( buffer, len, state ); \
\
for( j = 0; j <= len - 4; j += 4 ) \
{ \
worktype f0, f1; \
\
f0 = cast_macro1( buffer[j]*p[j+12] + p[j] ); \
f1 = cast_macro1( buffer[j+1]*p[j+13] + p[j+1] ); \
arr[i+j] = cast_macro2(f0); \
arr[i+j+1] = cast_macro2(f1); \
\
f0 = cast_macro1( buffer[j+2]*p[j+14] + p[j+2] ); \
f1 = cast_macro1( buffer[j+3]*p[j+15] + p[j+3] ); \
arr[i+j+2] = cast_macro2(f0); \
arr[i+j+3] = cast_macro2(f1); \
\
if( --k == 0 ) \
{ \
k = 3; \
p -= 12; \
} \
} \
\
for( ; j < len; j++ ) \
{ \
worktype f0 = cast_macro1( buffer[j]*p[j+12] + p[j] ); \
arr[i+j] = cast_macro2(f0); \
} \
} \
} \
\
return CV_OK; \
}
ICV_IMPL_RAND_BITS( 8u, uchar, CV_CAST_8U )
ICV_IMPL_RAND_BITS( 16u, ushort, CV_CAST_16U )
ICV_IMPL_RAND_BITS( 16s, short, CV_CAST_16S )
ICV_IMPL_RAND_BITS( 32s, int, CV_CAST_32S )
ICV_IMPL_RAND( 8u, uchar, int, cvFloor, CV_CAST_8U )
ICV_IMPL_RAND( 16u, ushort, int, cvFloor, CV_CAST_16U )
ICV_IMPL_RAND( 16s, short, int, cvFloor, CV_CAST_16S )
ICV_IMPL_RAND( 32s, int, int, cvFloor, CV_CAST_32S )
ICV_IMPL_RAND( 32f, float, float, CV_CAST_32F, CV_NOP )
ICV_IMPL_RANDN( 8u, uchar, int, cvRound, CV_CAST_8U )
ICV_IMPL_RANDN( 16u, ushort, int, cvRound, CV_CAST_16U )
ICV_IMPL_RANDN( 16s, short, int, cvRound, CV_CAST_16S )
ICV_IMPL_RANDN( 32s, int, int, cvRound, CV_CAST_32S )
ICV_IMPL_RANDN( 32f, float, float, CV_CAST_32F, CV_NOP )
ICV_IMPL_RANDN( 64f, double, double, CV_CAST_64F, CV_NOP )
static void icvInitRandTable( CvFuncTable* fastrng_tab,
CvFuncTable* rng_tab,
CvFuncTable* normal_tab )
{
fastrng_tab->fn_2d[CV_8U] = (void*)icvRandBits_8u_C1R;
fastrng_tab->fn_2d[CV_8S] = 0;
fastrng_tab->fn_2d[CV_16U] = (void*)icvRandBits_16u_C1R;
fastrng_tab->fn_2d[CV_16S] = (void*)icvRandBits_16s_C1R;
fastrng_tab->fn_2d[CV_32S] = (void*)icvRandBits_32s_C1R;
rng_tab->fn_2d[CV_8U] = (void*)icvRand_8u_C1R;
rng_tab->fn_2d[CV_8S] = 0;
rng_tab->fn_2d[CV_16U] = (void*)icvRand_16u_C1R;
rng_tab->fn_2d[CV_16S] = (void*)icvRand_16s_C1R;
rng_tab->fn_2d[CV_32S] = (void*)icvRand_32s_C1R;
rng_tab->fn_2d[CV_32F] = (void*)icvRand_32f_C1R;
rng_tab->fn_2d[CV_64F] = (void*)icvRand_64f_C1R;
normal_tab->fn_2d[CV_8U] = (void*)icvRandn_8u_C1R;
normal_tab->fn_2d[CV_8S] = 0;
normal_tab->fn_2d[CV_16U] = (void*)icvRandn_16u_C1R;
normal_tab->fn_2d[CV_16S] = (void*)icvRandn_16s_C1R;
normal_tab->fn_2d[CV_32S] = (void*)icvRandn_32s_C1R;
normal_tab->fn_2d[CV_32F] = (void*)icvRandn_32f_C1R;
normal_tab->fn_2d[CV_64F] = (void*)icvRandn_64f_C1R;
}
CV_IMPL void
cvRandArr( CvRNG* rng, CvArr* arr, int disttype, CvScalar param1, CvScalar param2 )
{
static CvFuncTable rng_tab[2], fastrng_tab;
static int inittab = 0;
CV_FUNCNAME( "cvRandArr" );
__BEGIN__;
int is_nd = 0;
CvMat stub, *mat = (CvMat*)arr;
int type, depth, channels;
double dparam[2][12];
int iparam[2][12];
void* param = dparam;
int i, fast_int_mode = 0;
int mat_step = 0;
CvSize size;
CvFunc2D_1A2P func = 0;
CvMatND stub_nd;
CvNArrayIterator iterator_state, *iterator = 0;
if( !inittab )
{
icvInitRandTable( &fastrng_tab, &rng_tab[CV_RAND_UNI],
&rng_tab[CV_RAND_NORMAL] );
inittab = 1;
}
if( !rng )
CV_ERROR( CV_StsNullPtr, "Null pointer to RNG state" );
if( CV_IS_MATND(mat) )
{
iterator = &iterator_state;
CV_CALL( cvInitNArrayIterator( 1, &arr, 0, &stub_nd, iterator ));
type = CV_MAT_TYPE(iterator->hdr[0]->type);
size = iterator->size;
is_nd = 1;
}
else
{
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" );
}
type = CV_MAT_TYPE( mat->type );
size = cvGetMatSize( mat );
mat_step = mat->step;
if( mat->height > 1 && CV_IS_MAT_CONT( mat->type ))
{
size.width *= size.height;
mat_step = CV_STUB_STEP;
size.height = 1;
}
}
depth = CV_MAT_DEPTH( type );
channels = CV_MAT_CN( type );
size.width *= channels;
if( disttype == CV_RAND_UNI )
{
if( depth <= CV_32S )
{
for( i = 0, fast_int_mode = 1; i < channels; i++ )
{
int t0 = iparam[0][i] = cvCeil( param1.val[i] );
int t1 = iparam[1][i] = cvFloor( param2.val[i] ) - t0;
double diff = param1.val[i] - param2.val[i];
fast_int_mode &= INT_MIN <= diff && diff <= INT_MAX && (t1 & (t1 - 1)) == 0;
}
}
if( fast_int_mode )
{
for( i = 0; i < channels; i++ )
iparam[1][i]--;
for( ; i < 12; i++ )
{
int t0 = iparam[0][i - channels];
int t1 = iparam[1][i - channels];
iparam[0][i] = t0;
iparam[1][i] = t1;
}
CV_GET_FUNC_PTR( func, (CvFunc2D_1A2P)(fastrng_tab.fn_2d[depth]));
param = iparam;
}
else
{
for( i = 0; i < channels; i++ )
{
double t0 = param1.val[i];
double t1 = param2.val[i];
dparam[0][i] = t0 - (t1 - t0);
dparam[1][i] = t1 - t0;
}
CV_GET_FUNC_PTR( func, (CvFunc2D_1A2P)(rng_tab[0].fn_2d[depth]));
}
}
else if( disttype == CV_RAND_NORMAL )
{
for( i = 0; i < channels; i++ )
{
double t0 = param1.val[i];
double t1 = param2.val[i];
dparam[0][i] = t0;
dparam[1][i] = t1;
}
CV_GET_FUNC_PTR( func, (CvFunc2D_1A2P)(rng_tab[1].fn_2d[depth]));
}
else
{
CV_ERROR( CV_StsBadArg, "Unknown distribution type" );
}
if( !fast_int_mode )
{
for( i = channels; i < 12; i++ )
{
double t0 = dparam[0][i - channels];
double t1 = dparam[1][i - channels];
dparam[0][i] = t0;
dparam[1][i] = t1;
}
}
if( !is_nd )
{
IPPI_CALL( func( mat->data.ptr, mat_step, size, rng, param ));
}
else
{
do
{
IPPI_CALL( func( iterator->ptr[0], CV_STUB_STEP, size, rng, param ));
}
while( cvNextNArraySlice( iterator ));
}
__END__;
}
/* End of file. */