Public Member Functions | |
std::size_t | getJacSize () const |
Get the number of Jacobian images (= state dimension). | |
std::size_t | getMeasSize () const |
Get the actual measurement size = size of the ROI. | |
bool | getUseScalarResidual () const |
Return flag for computing scalar image residual. | |
void | resizeAll (std::size_t width, std::size_t height, opentl::core::cvdata::Image::ColorChannels channels=opentl::core::cvdata::Image::RGB, opentl::core::cvdata::Image::BitsPerChannel bpc=opentl::core::cvdata::Image::BPC8U, std::size_t jacSize_=0) |
Set image sizes, number of channels and bits per channel (the same for all data), and Jacobian dimension. | |
void | setUseScalarResidual (bool val) |
Set flag for computing scalar image residual. | |
T_MEAS_PIX (bool scalarResidual=true) | |
Public Attributes | |
boost::shared_ptr < opentl::core::cvdata::Image > | E_image |
E_image = global image residual. Same number of channels of z and h, but internal type 64-bit floating point (double). | |
double | E_scalar |
E = scalar, global image residual between z and h. | |
boost::shared_ptr < opentl::core::cvdata::Image > | E_valid |
E_valid = pixel-wise validity mask for image residuals (0=missing, 1=valid). 1-channel, unsigned char. | |
boost::shared_ptr < opentl::core::cvdata::Image > | h |
h = expected pixel responses (e.g. target shadows). It has the same channels and internal type of z. | |
double | invCov |
invCov = inverse covariance of E (scalar, the same for each pixel) | |
std::vector< boost::shared_ptr < opentl::core::cvdata::Image > > | Jac_image |
Jac = pixel-wise Jacobian of h (set of derivative images). Each one has the same channels and type of E_image (floating point). | |
std::vector< double > | Jac_scalar |
Jac = Jacobian of scalar residual. | |
boost::shared_ptr < opentl::core::cvdata::Image > | z |
z = measured pixel responses. It may have an arbitrary number of channel and internal type. |
opentl::core::cvdata::T_MEAS_PIX::T_MEAS_PIX | ( | bool | scalarResidual = true |
) | [inline] |
std::size_t opentl::core::cvdata::T_MEAS_PIX::getJacSize | ( | ) | const [inline] |
Get the number of Jacobian images (= state dimension).
std::size_t opentl::core::cvdata::T_MEAS_PIX::getMeasSize | ( | ) | const [inline] |
Get the actual measurement size = size of the ROI.
bool opentl::core::cvdata::T_MEAS_PIX::getUseScalarResidual | ( | ) | const [inline] |
Return flag for computing scalar image residual.
void opentl::core::cvdata::T_MEAS_PIX::resizeAll | ( | std::size_t | width, | |
std::size_t | height, | |||
opentl::core::cvdata::Image::ColorChannels | channels = opentl::core::cvdata::Image::RGB , |
|||
opentl::core::cvdata::Image::BitsPerChannel | bpc = opentl::core::cvdata::Image::BPC8U , |
|||
std::size_t | jacSize_ = 0 | |||
) | [inline] |
Set image sizes, number of channels and bits per channel (the same for all data), and Jacobian dimension.
void opentl::core::cvdata::T_MEAS_PIX::setUseScalarResidual | ( | bool | val | ) | [inline] |
Set flag for computing scalar image residual.
boost::shared_ptr<opentl::core::cvdata::Image> opentl::core::cvdata::T_MEAS_PIX::E_image |
E_image = global image residual. Same number of channels of z and h, but internal type 64-bit floating point (double).
E = scalar, global image residual between z and h.
boost::shared_ptr<opentl::core::cvdata::Image> opentl::core::cvdata::T_MEAS_PIX::E_valid |
E_valid = pixel-wise validity mask for image residuals (0=missing, 1=valid). 1-channel, unsigned char.
boost::shared_ptr<opentl::core::cvdata::Image> opentl::core::cvdata::T_MEAS_PIX::h |
h = expected pixel responses (e.g. target shadows). It has the same channels and internal type of z.
invCov = inverse covariance of E (scalar, the same for each pixel)
std::vector<boost::shared_ptr<opentl::core::cvdata::Image> > opentl::core::cvdata::T_MEAS_PIX::Jac_image |
Jac = pixel-wise Jacobian of h (set of derivative images). Each one has the same channels and type of E_image (floating point).
std::vector<double> opentl::core::cvdata::T_MEAS_PIX::Jac_scalar |
Jac = Jacobian of scalar residual.
boost::shared_ptr<opentl::core::cvdata::Image> opentl::core::cvdata::T_MEAS_PIX::z |
z = measured pixel responses. It may have an arbitrary number of channel and internal type.