513 lines
11 KiB
C++
513 lines
11 KiB
C++
|
#include <math.h>
|
||
|
#include "osg/Matrix"
|
||
|
#include "osg/Input"
|
||
|
#include "osg/Output"
|
||
|
#include "osg/Notify"
|
||
|
|
||
|
#define square(x) ((x)*(x))
|
||
|
#define DEG2RAD(x) ((x)*M_PI/180.0)
|
||
|
|
||
|
using namespace osg;
|
||
|
|
||
|
typedef struct quaternion_
|
||
|
{
|
||
|
double x ;
|
||
|
double y ;
|
||
|
double z ;
|
||
|
double w ;
|
||
|
} quaternion ;
|
||
|
|
||
|
/* C = a(row).b(row) */
|
||
|
|
||
|
#define matrix_inner_product( a, b, row, col, C ) \
|
||
|
{ \
|
||
|
(C)[row][col] = (a)[row][0] * (b)[0][col] + \
|
||
|
(a)[row][1] * (b)[1][col] + \
|
||
|
(a)[row][2] * (b)[2][col] + \
|
||
|
(a)[row][3] * (b)[3][col]; \
|
||
|
}
|
||
|
|
||
|
/* C = a.b */
|
||
|
|
||
|
#define matrix_mult( a, b, C ) \
|
||
|
{ \
|
||
|
matrix_inner_product( a, b, 0, 0, C ); \
|
||
|
matrix_inner_product( a, b, 0, 1, C ); \
|
||
|
matrix_inner_product( a, b, 0, 2, C ); \
|
||
|
matrix_inner_product( a, b, 0, 3, C ); \
|
||
|
matrix_inner_product( a, b, 1, 0, C ); \
|
||
|
matrix_inner_product( a, b, 1, 1, C ); \
|
||
|
matrix_inner_product( a, b, 1, 2, C ); \
|
||
|
matrix_inner_product( a, b, 1, 3, C ); \
|
||
|
matrix_inner_product( a, b, 2, 0, C ); \
|
||
|
matrix_inner_product( a, b, 2, 1, C ); \
|
||
|
matrix_inner_product( a, b, 2, 2, C ); \
|
||
|
matrix_inner_product( a, b, 2, 3, C ); \
|
||
|
matrix_inner_product( a, b, 3, 0, C ); \
|
||
|
matrix_inner_product( a, b, 3, 1, C ); \
|
||
|
matrix_inner_product( a, b, 3, 2, C ); \
|
||
|
matrix_inner_product( a, b, 3, 3, C ); \
|
||
|
}
|
||
|
|
||
|
static void quaternion_matrix( quaternion *q, double mat[4][4] )
|
||
|
{
|
||
|
/* copied from Shoemake/ACM SIGGRAPH 89 */
|
||
|
double xs, ys, zs, wx, wy, wz, xx, xy, xz, yy, yz, zz ;
|
||
|
|
||
|
xs = q->x + q->x;
|
||
|
ys = q->y + q->y;
|
||
|
zs = q->z + q->z;
|
||
|
|
||
|
wx = q->w * xs ; wy = q->w * ys ; wz = q->w * zs ;
|
||
|
xx = q->x * xs ; xy = q->x * ys ; xz = q->x * zs ;
|
||
|
yy = q->y * ys ; yz = q->y * zs ; zz = q->z * zs ;
|
||
|
|
||
|
mat[0][0] = 1.0 - ( yy + zz ) ;
|
||
|
mat[0][1] = xy - wz ;
|
||
|
mat[0][2] = xz + wy ;
|
||
|
mat[1][0] = xy + wz ;
|
||
|
mat[1][1] = 1.0 - ( xx + zz ) ;
|
||
|
mat[1][2] = yz - wx ;
|
||
|
mat[2][0] = xz - wy ;
|
||
|
mat[2][1] = yz + wx ;
|
||
|
mat[2][2] = 1.0 - ( xx + yy ) ;
|
||
|
|
||
|
mat[0][3] = 0.0;
|
||
|
mat[1][3] = 0.0;
|
||
|
mat[2][3] = 0.0;
|
||
|
|
||
|
mat[3][0] = 0.0;
|
||
|
mat[3][1] = 0.0;
|
||
|
mat[3][2] = 0.0;
|
||
|
mat[3][3] = 1.0;
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix::Matrix()
|
||
|
{
|
||
|
makeIdent();
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix::Matrix(const Matrix& matrix) : Object()
|
||
|
{
|
||
|
memcpy(_mat,matrix._mat,sizeof(_mat));
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix& Matrix::operator = (const Matrix& matrix)
|
||
|
{
|
||
|
if (&matrix==this) return *this;
|
||
|
memcpy(_mat,matrix._mat,sizeof(_mat));
|
||
|
return *this;
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix::Matrix(
|
||
|
float a00, float a01, float a02, float a03,
|
||
|
float a10, float a11, float a12, float a13,
|
||
|
float a20, float a21, float a22, float a23,
|
||
|
float a30, float a31, float a32, float a33)
|
||
|
{
|
||
|
_mat[0][0] = a00;
|
||
|
_mat[0][1] = a01;
|
||
|
_mat[0][2] = a02;
|
||
|
_mat[0][3] = a03;
|
||
|
|
||
|
_mat[1][0] = a10;
|
||
|
_mat[1][1] = a11;
|
||
|
_mat[1][2] = a12;
|
||
|
_mat[1][3] = a13;
|
||
|
|
||
|
_mat[2][0] = a20;
|
||
|
_mat[2][1] = a21;
|
||
|
_mat[2][2] = a22;
|
||
|
_mat[2][3] = a23;
|
||
|
|
||
|
_mat[3][0] = a30;
|
||
|
_mat[3][1] = a31;
|
||
|
_mat[3][2] = a32;
|
||
|
_mat[3][3] = a33;
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix::~Matrix()
|
||
|
{
|
||
|
}
|
||
|
|
||
|
|
||
|
Matrix* Matrix::instance()
|
||
|
{
|
||
|
static ref_ptr<Matrix> s_matrix(new Matrix());
|
||
|
return s_matrix.get();
|
||
|
}
|
||
|
|
||
|
|
||
|
bool Matrix::readLocalData(Input& fr)
|
||
|
{
|
||
|
bool iteratorAdvanced = false;
|
||
|
bool matched = true;
|
||
|
for(int k=0;k<16 && matched;++k)
|
||
|
{
|
||
|
matched = fr[k].isFloat();
|
||
|
}
|
||
|
if (matched)
|
||
|
{
|
||
|
int k=0;
|
||
|
for(int i=0;i<4;++i)
|
||
|
{
|
||
|
for(int j=0;j<4;++j)
|
||
|
{
|
||
|
fr[k].getFloat(_mat[i][j]);
|
||
|
k++;
|
||
|
}
|
||
|
}
|
||
|
fr += 16;
|
||
|
iteratorAdvanced = true;
|
||
|
}
|
||
|
|
||
|
return iteratorAdvanced;
|
||
|
}
|
||
|
|
||
|
|
||
|
bool Matrix::writeLocalData(Output& fw)
|
||
|
{
|
||
|
fw.indent() << _mat[0][0] << " " << _mat[0][1] << " " << _mat[0][2] << " " << _mat[0][3] << endl;
|
||
|
fw.indent() << _mat[1][0] << " " << _mat[1][1] << " " << _mat[1][2] << " " << _mat[1][3] << endl;
|
||
|
fw.indent() << _mat[2][0] << " " << _mat[2][1] << " " << _mat[2][2] << " " << _mat[2][3] << endl;
|
||
|
fw.indent() << _mat[3][0] << " " << _mat[3][1] << " " << _mat[3][2] << " " << _mat[3][3] << endl;
|
||
|
return true;
|
||
|
}
|
||
|
|
||
|
|
||
|
void Matrix::makeIdent()
|
||
|
{
|
||
|
_mat[0][0] = 1.0f;
|
||
|
_mat[0][1] = 0.0f;
|
||
|
_mat[0][2] = 0.0f;
|
||
|
_mat[0][3] = 0.0f;
|
||
|
|
||
|
_mat[1][0] = 0.0f;
|
||
|
_mat[1][1] = 1.0f;
|
||
|
_mat[1][2] = 0.0f;
|
||
|
_mat[1][3] = 0.0f;
|
||
|
|
||
|
_mat[2][0] = 0.0f;
|
||
|
_mat[2][1] = 0.0f;
|
||
|
_mat[2][2] = 1.0f;
|
||
|
_mat[2][3] = 0.0f;
|
||
|
|
||
|
_mat[3][0] = 0.0f;
|
||
|
_mat[3][1] = 0.0f;
|
||
|
_mat[3][2] = 0.0f;
|
||
|
_mat[3][3] = 1.0f;
|
||
|
}
|
||
|
|
||
|
void Matrix::copy(const Matrix& matrix)
|
||
|
{
|
||
|
memcpy(_mat,matrix._mat,sizeof(_mat));
|
||
|
}
|
||
|
|
||
|
void Matrix::makeScale(float sx, float sy, float sz)
|
||
|
{
|
||
|
makeIdent();
|
||
|
_mat[0][0] = sx;
|
||
|
_mat[1][1] = sy;
|
||
|
_mat[2][2] = sz;
|
||
|
}
|
||
|
|
||
|
|
||
|
void Matrix::preScale( float sx, float sy, float sz, const Matrix& m )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeScale(sx, sy, sz);
|
||
|
mult(transMat,m);
|
||
|
}
|
||
|
|
||
|
void Matrix::postScale( const Matrix& m, float sx, float sy, float sz )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeScale(sx, sy, sz);
|
||
|
mult(m,transMat);
|
||
|
}
|
||
|
|
||
|
void Matrix::preScale( float sx, float sy, float sz )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeScale(sx, sy, sz);
|
||
|
preMult(transMat);
|
||
|
}
|
||
|
|
||
|
void Matrix::postScale( float sx, float sy, float sz )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeScale(sx, sy, sz);
|
||
|
postMult(transMat);
|
||
|
}
|
||
|
|
||
|
|
||
|
|
||
|
|
||
|
void Matrix::makeTrans( float tx, float ty, float tz )
|
||
|
{
|
||
|
makeIdent();
|
||
|
_mat[3][0] = tx;
|
||
|
_mat[3][1] = ty;
|
||
|
_mat[3][2] = tz;
|
||
|
}
|
||
|
|
||
|
void Matrix::preTrans( float tx, float ty, float tz, const Matrix& m )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeTrans(tx, ty, tz);
|
||
|
mult(transMat,m);
|
||
|
}
|
||
|
|
||
|
void Matrix::postTrans( const Matrix& m, float tx, float ty, float tz )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeTrans(tx, ty, tz);
|
||
|
mult(m,transMat);
|
||
|
}
|
||
|
|
||
|
void Matrix::preTrans( float tx, float ty, float tz )
|
||
|
{
|
||
|
_mat[3][0] = (tx * _mat[0][0]) + (ty * _mat[1][0]) + (tz * _mat[2][0]) + _mat[3][0];
|
||
|
_mat[3][1] = (tx * _mat[0][1]) + (ty * _mat[1][1]) + (tz * _mat[2][1]) + _mat[3][1];
|
||
|
_mat[3][2] = (tx * _mat[0][2]) + (ty * _mat[1][2]) + (tz * _mat[2][2]) + _mat[3][2];
|
||
|
_mat[3][3] = (tx * _mat[0][3]) + (ty * _mat[1][3]) + (tz * _mat[2][3]) + _mat[3][3];
|
||
|
}
|
||
|
|
||
|
void Matrix::postTrans( float tx, float ty, float tz )
|
||
|
{
|
||
|
Matrix transMat;
|
||
|
transMat.makeTrans(tx, ty, tz);
|
||
|
postMult(transMat);
|
||
|
}
|
||
|
|
||
|
void Matrix::makeRot( float deg, float x, float y, float z )
|
||
|
{
|
||
|
double __mat[4][4];
|
||
|
quaternion q;
|
||
|
float d = sqrtf( square(x) + square(y) + square(z) );
|
||
|
|
||
|
if( d == 0 )
|
||
|
return;
|
||
|
|
||
|
float sin_HalfAngle = sinf( DEG2RAD(deg/2) );
|
||
|
float cos_HalfAngle = cosf( DEG2RAD(deg/2) );
|
||
|
|
||
|
q.x = sin_HalfAngle * (x/d);
|
||
|
q.y = sin_HalfAngle * (y/d);
|
||
|
q.z = sin_HalfAngle * (z/d);
|
||
|
q.w = cos_HalfAngle;
|
||
|
|
||
|
quaternion_matrix( &q, __mat );
|
||
|
|
||
|
for(int i=0;i<4;++i)
|
||
|
{
|
||
|
for(int j=0;j<4;++j)
|
||
|
{
|
||
|
_mat[i][j]=__mat[i][j];
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
void Matrix::preRot( float deg, float x, float y, float z, const Matrix& m )
|
||
|
{
|
||
|
Matrix rotMat;
|
||
|
rotMat.makeRot( deg, x, y, z );
|
||
|
mult(rotMat,m);
|
||
|
}
|
||
|
|
||
|
void Matrix::postRot( const Matrix& m, float deg, float x, float y, float z )
|
||
|
{
|
||
|
Matrix rotMat;
|
||
|
rotMat.makeRot( deg, x, y, z );
|
||
|
mult(m,rotMat);
|
||
|
}
|
||
|
|
||
|
void Matrix::preRot( float deg, float x, float y, float z )
|
||
|
{
|
||
|
quaternion q;
|
||
|
double __mat[4][4];
|
||
|
float res_mat[4][4];
|
||
|
|
||
|
float d = sqrtf( square(x) + square(y) + square(z) );
|
||
|
|
||
|
if( d == 0 )
|
||
|
return;
|
||
|
|
||
|
float sin_HalfAngle = sinf( DEG2RAD(deg/2) );
|
||
|
float cos_HalfAngle = cosf( DEG2RAD(deg/2) );
|
||
|
|
||
|
q.x = sin_HalfAngle * (x/d);
|
||
|
q.y = sin_HalfAngle * (y/d);
|
||
|
q.z = sin_HalfAngle * (z/d);
|
||
|
q.w = cos_HalfAngle;
|
||
|
|
||
|
quaternion_matrix( &q, __mat );
|
||
|
matrix_mult( __mat, _mat, res_mat );
|
||
|
memcpy( _mat, res_mat, sizeof( _mat ) );
|
||
|
}
|
||
|
|
||
|
void Matrix::postRot( float deg, float x, float y, float z )
|
||
|
{
|
||
|
quaternion q;
|
||
|
double __mat[4][4];
|
||
|
float res_mat[4][4];
|
||
|
|
||
|
float d = sqrtf( square(x) + square(y) + square(z) );
|
||
|
|
||
|
if( d == 0 )
|
||
|
return;
|
||
|
|
||
|
float sin_HalfAngle = sinf( DEG2RAD(deg/2) );
|
||
|
float cos_HalfAngle = cosf( DEG2RAD(deg/2) );
|
||
|
|
||
|
q.x = sin_HalfAngle * (x/d);
|
||
|
q.y = sin_HalfAngle * (y/d);
|
||
|
q.z = sin_HalfAngle * (z/d);
|
||
|
q.w = cos_HalfAngle;
|
||
|
|
||
|
quaternion_matrix( &q, __mat );
|
||
|
matrix_mult( _mat, __mat , res_mat );
|
||
|
memcpy( _mat, res_mat, sizeof( _mat ) );
|
||
|
}
|
||
|
|
||
|
void Matrix::setTrans( float tx, float ty, float tz )
|
||
|
{
|
||
|
_mat[3][0] = tx;
|
||
|
_mat[3][1] = ty;
|
||
|
_mat[3][2] = tz;
|
||
|
}
|
||
|
void Matrix::setTrans( const Vec3& v )
|
||
|
{
|
||
|
_mat[3][0] = v[0];
|
||
|
_mat[3][1] = v[1];
|
||
|
_mat[3][2] = v[2];
|
||
|
}
|
||
|
|
||
|
void Matrix::preMult(const Matrix& m)
|
||
|
{
|
||
|
Matrix tm;
|
||
|
matrix_mult( m._mat, _mat, tm._mat );
|
||
|
*this = tm;
|
||
|
}
|
||
|
|
||
|
void Matrix::postMult(const Matrix& m)
|
||
|
{
|
||
|
Matrix tm;
|
||
|
matrix_mult( _mat, m._mat, tm._mat );
|
||
|
*this = tm;
|
||
|
}
|
||
|
|
||
|
void Matrix::mult(const Matrix& lhs,const Matrix& rhs)
|
||
|
{
|
||
|
matrix_mult( lhs._mat, rhs._mat, _mat );
|
||
|
}
|
||
|
|
||
|
Matrix Matrix::operator * (const Matrix& m) const
|
||
|
{
|
||
|
Matrix nm;
|
||
|
matrix_mult( _mat,m._mat, nm._mat );
|
||
|
return nm;
|
||
|
}
|
||
|
|
||
|
|
||
|
bool Matrix::invert(const Matrix& _m)
|
||
|
{
|
||
|
// code lifted from VR Juggler.
|
||
|
// not cleanly added, but seems to work. RO.
|
||
|
|
||
|
const float* a = reinterpret_cast<const float*>(_m._mat);
|
||
|
float* b = reinterpret_cast<float*>(_mat);
|
||
|
|
||
|
int n = 4;
|
||
|
int i, j, k;
|
||
|
int r[ 4], c[ 4], row[ 4], col[ 4];
|
||
|
float m[ 4][ 4*2], pivot, max_m, tmp_m, fac;
|
||
|
|
||
|
/* Initialization */
|
||
|
for ( i = 0; i < n; i ++ )
|
||
|
{
|
||
|
r[ i] = c[ i] = 0;
|
||
|
row[ i] = col[ i] = 0;
|
||
|
}
|
||
|
|
||
|
/* Set working matrix */
|
||
|
for ( i = 0; i < n; i++ )
|
||
|
{
|
||
|
for ( j = 0; j < n; j++ )
|
||
|
{
|
||
|
m[ i][ j] = a[ i * n + j];
|
||
|
m[ i][ j + n] = ( i == j ) ? 1.0 : 0.0 ;
|
||
|
}
|
||
|
}
|
||
|
|
||
|
/* Begin of loop */
|
||
|
for ( k = 0; k < n; k++ )
|
||
|
{
|
||
|
/* Choosing the pivot */
|
||
|
for ( i = 0, max_m = 0; i < n; i++ )
|
||
|
{
|
||
|
if ( row[ i] ) continue;
|
||
|
for ( j = 0; j < n; j++ )
|
||
|
{
|
||
|
if ( col[ j] ) continue;
|
||
|
tmp_m = fabs( m[ i][j]);
|
||
|
if ( tmp_m > max_m)
|
||
|
{
|
||
|
max_m = tmp_m;
|
||
|
r[ k] = i;
|
||
|
c[ k] = j;
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
row[ r[k] ] = col[ c[k] ] = 1;
|
||
|
pivot = m[ r[ k] ][ c[ k] ];
|
||
|
|
||
|
if ( fabs( pivot) <= 1e-20)
|
||
|
{
|
||
|
notify(WARN) << "*** pivot = %f in mat_inv. ***\n";
|
||
|
//exit( 0);
|
||
|
return false;
|
||
|
}
|
||
|
|
||
|
/* Normalization */
|
||
|
for ( j = 0; j < 2*n; j++ )
|
||
|
{
|
||
|
if ( j == c[ k] )
|
||
|
m[ r[ k]][ j] = 1.0;
|
||
|
else
|
||
|
m[ r[ k]][ j] /=pivot;
|
||
|
}
|
||
|
|
||
|
/* Reduction */
|
||
|
for ( i = 0; i < n; i++ )
|
||
|
{
|
||
|
if ( i == r[ k] )
|
||
|
continue;
|
||
|
|
||
|
for ( j=0, fac = m[ i][ c[k]];j < 2*n; j++ )
|
||
|
{
|
||
|
if ( j == c[ k] )
|
||
|
m[ i][ j] =0.0;
|
||
|
else
|
||
|
m[ i][ j] -=fac * m[ r[k]][ j];
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
|
||
|
/* Assign invers to a matrix */
|
||
|
for ( i = 0; i < n; i++ )
|
||
|
for ( j = 0; j < n; j++ )
|
||
|
row[ i] = ( c[ j] == i ) ? r[j] : row[ i];
|
||
|
|
||
|
for ( i = 0; i < n; i++ )
|
||
|
for ( j = 0; j < n; j++ )
|
||
|
b[ i * n + j] = m[ row[ i]][j + n];
|
||
|
|
||
|
return true; // It worked
|
||
|
}
|