This menu allows selection of how particles which do not have a velocity history will be treated. Normally the velocity at the previous time step is used to predict the position of the particle at the new time step. If however there is no velocity history for a given particle at the previous time step, we must decide how we are going to predict its position at the new time step. Note that this velocity prediction will be subject to some error. The limits on this error are set by [qM Max velocity error for new paths].
Under this option DigImage assumes the velocity for new particles is, on average, constant in space and time. The x and y components of this mean new velocity are specified.
This option allows the user to specify a formula to give the predicted velocity for new particle paths. The formula may be given in terms of space (world coordinates X,Y or pixel coordinates I,J) and time (T).
This option is only appropriate with relatively thin light sheets or in flows which are approximately two dimensional. It evaluates a spatially varying local estimate for the velocity utilising the velocities of particles for which a match has been found for at least two time steps. This estimate is then used as the guess for the new particle velocity to be used at the next time step. Note that if there is insufficient data to obtain such an estimate (eg. on the first time step where there is no velocity history), then a zero velocity is assumed. [M Local velocity estimate with constant if too few particles] offers a more general alternative to this option.
This option is a combination of [C Constant predicted velocity] and [Local velocity estimate]. If there are sufficiently many particles for which the velocity is known in the neighbourhood of the new particle, then the local mean velocity is used to estimate the new particle velocity. However, if there are no particles with known velocity, then the specified constant velocity is used. For between one and five particles, a weighted mean of the two limits is employed.
This option represents a cross between [L Local velocity estimate] and [R Region-dependent predicted velocity]. If there are sufficient particles with valid matches for a local velocity estimate to be used, then it will be, otherwise a formula in space and time is used to predict the velocity.
This option represents a cross between [L Local velocity estimate] and [R Region-dependent predicted velocity]. If there are sufficient particles with valid matches for a local velocity estimate to be used, then it will be, otherwise a region-dependent velocity is used.
This option is designed for use with the [qCRR Record which region particle in] contour mask option. It allows specification of two separate constant velocities. Which one is used by a given particle depends on whether the particle falls in the low or high intensity region.
This option is effectively a subset of [C Constant predicted velocity]. It assumes that the mean new velocity is spatially and temporaly zero.