Browse Source

Math: change signs in quaternion library

thanks to Justin for the suggestion
master
Andrew Tridgell 13 years ago
parent
commit
d948a28e7c
  1. 20
      libraries/AP_Math/quaternion.cpp

20
libraries/AP_Math/quaternion.cpp

@ -34,12 +34,12 @@ void Quaternion::rotation_matrix(Matrix3f &m) @@ -34,12 +34,12 @@ void Quaternion::rotation_matrix(Matrix3f &m)
m.a.x = 1-2*(q3q3 + q4q4);
m.a.y = 2*(q2q3 - q1q4);
m.a.z = - 2*(q2q4 + q1q3);
m.a.z = 2*(q2q4 + q1q3);
m.b.x = 2*(q2q3 + q1q4);
m.b.y = 1-2*(q2q2 + q4q4);
m.b.z = -2*(q3q4 - q1q2);
m.c.x = -2*(q2q4 - q1q3);
m.c.y = -2*(q3q4 + q1q2);
m.b.z = 2*(q3q4 - q1q2);
m.c.x = 2*(q2q4 - q1q3);
m.c.y = 2*(q3q4 + q1q2);
m.c.z = 1-2*(q2q2 + q3q3);
}
@ -61,10 +61,8 @@ void Quaternion::from_euler(float roll, float pitch, float yaw) @@ -61,10 +61,8 @@ void Quaternion::from_euler(float roll, float pitch, float yaw)
float cr2 = cos(roll*0.5);
float cp2 = cos(pitch*0.5);
float cy2 = cos(yaw*0.5);
// the sign reversal here is due to the different conventions
// in the madgwick quaternion code and the rest of APM
float sr2 = -sin(roll*0.5);
float sp2 = -sin(pitch*0.5);
float sr2 = sin(roll*0.5);
float sp2 = sin(pitch*0.5);
float sy2 = sin(yaw*0.5);
q1 = cr2*cp2*cy2 + sr2*sp2*sy2;
@ -77,12 +75,12 @@ void Quaternion::from_euler(float roll, float pitch, float yaw) @@ -77,12 +75,12 @@ void Quaternion::from_euler(float roll, float pitch, float yaw)
void Quaternion::to_euler(float *roll, float *pitch, float *yaw)
{
if (roll) {
*roll = -(atan2(2.0*(q1*q2 + q3*q4),
1 - 2.0*(q2*q2 + q3*q3)));
*roll = (atan2(2.0*(q1*q2 + q3*q4),
1 - 2.0*(q2*q2 + q3*q3)));
}
if (pitch) {
// we let safe_asin() handle the singularities near 90/-90 in pitch
*pitch = -safe_asin(2.0*(q1*q3 - q4*q2));
*pitch = safe_asin(2.0*(q1*q3 - q4*q2));
}
if (yaw) {
*yaw = atan2(2.0*(q1*q4 + q2*q3),

Loading…
Cancel
Save