Detailed Description
Instantaneous state uncertainty estimates.
Data in these fields represent the uncertainties of the corresponding fields in dwEgomotionResult. The uncertainties are represented as covariance matrix when appropriate, or in standard deviation around the estimate otherwise. These values are provided "as is" for usage in fusion algorithms depending on egomotion estimation but no accuracy claims are made (see below).
- Note
- The uncertainties do not represent estimate accuracy, only the precision of the estimate, under the classical filtering assumptions and limitation of the state and models used in internal estimators:
- error distributions are zero mean
- error distributions are Gaussian Neither exactly hold in practice, at least not in all cases. Errors that do not fit the zero mean Gaussian distribution might include calibration errors, model inaccuracies and limitations, input signal degradations or outlier measurements. Such errors are unobservable (or the estimator would correct for them) and result in biased error distributions. Covariances are therefore not a suitable signal to assess egomotion accuracy (i.e. “quality”), particularly in safety-relevant context. Use dwEgomotionResultQuality instead.
- Units are identical to those used in dwEgomotionResult. Units in covariance matrices are squared.
- Rotation is represented as a quaternion; however, here a 3x3 covariance of the equivalent euler angles is given (order: roll, pitch, yaw) in [rad].
Definition at line 495 of file EgomotionTypes.h.
Data Fields | |
| dwMatrix3f | rotation |
| Rotation covariance represented as euler angles (order: roll, pitch, yaw) in [rad^2]. More... | |
| float32_t | linearVelocity [3] |
| Linear velocity std dev in body frame measured in [m/s]. More... | |
| float32_t | angularVelocity [3] |
| Rotation speed std dev in body frame measured in [rad/s]. More... | |
| float32_t | linearAcceleration [3] |
| Linear acceleration std dev measured in body frame in [m/s^2]. More... | |
| float32_t | angularAcceleration [3] |
| float32_t | curvature |
| Curvature std dev in body frame in [1/m]. More... | |
| dwTime_t | timestamp |
| Timestamp of egomotion uncertainty estimate [us]. More... | |
| int64_t | validFlags |
| Bitwise combination of dwEgomotionDataField flags. More... | |
Field Documentation
◆ angularAcceleration
| float32_t dwEgomotionUncertainty::angularAcceleration[3] |
Definition at line 501 of file EgomotionTypes.h.
◆ angularVelocity
| float32_t dwEgomotionUncertainty::angularVelocity[3] |
Rotation speed std dev in body frame measured in [rad/s].
Definition at line 499 of file EgomotionTypes.h.
◆ curvature
| float32_t dwEgomotionUncertainty::curvature |
Curvature std dev in body frame in [1/m].
Definition at line 502 of file EgomotionTypes.h.
◆ linearAcceleration
| float32_t dwEgomotionUncertainty::linearAcceleration[3] |
Linear acceleration std dev measured in body frame in [m/s^2].
Definition at line 500 of file EgomotionTypes.h.
◆ linearVelocity
| float32_t dwEgomotionUncertainty::linearVelocity[3] |
Linear velocity std dev in body frame measured in [m/s].
Definition at line 498 of file EgomotionTypes.h.
◆ rotation
| dwMatrix3f dwEgomotionUncertainty::rotation |
Rotation covariance represented as euler angles (order: roll, pitch, yaw) in [rad^2].
Definition at line 497 of file EgomotionTypes.h.
◆ timestamp
| dwTime_t dwEgomotionUncertainty::timestamp |
Timestamp of egomotion uncertainty estimate [us].
Definition at line 503 of file EgomotionTypes.h.
◆ validFlags
| int64_t dwEgomotionUncertainty::validFlags |
Bitwise combination of dwEgomotionDataField flags.
Definition at line 504 of file EgomotionTypes.h.
The documentation for this struct was generated from the following file: