Sethi and Jain SJ87

Original reference: [11].
Additional assumptions: No entry/exit. No occlusion in $F_{1}$, $F_{2}$, $F_{3}$.
Initialization: Correspondences between $F_{1}$ and $F_{2}$ are initially set using the nearest neighbors. These can be altered later.
Cost function:
 $$\begin{multline}\delta_{S}(P_{k-1,n},P_{k,i},P_{k+1,m}) = w_{1} \left( 1-\frac{... ...ert +\left\Vert \overline{% P_{k,i}P_{k+1,m}}\right\Vert }\right) \end{multline}$$
$\overline{P_{k-1,n}P_{k,i}}$ and $\overline{P_{k,i}P_{k+1,m}}$are the vectors pointing from $P_{k-1,n}$ to $P_{k,i}$ and from $P_{k,i}$ to $P_{k+1,m}$, respectively. The first term in (1) penalizes changes in the direction, the second in the magnitude of the speed vector. The weights are $w_{1}=0.1$, $w_{2}=0.9$. By default, the first term is set to zero if any of the two vectors is zero.
Linking strategy: Iterative greedy exchange in forward and backward directions.
Occlusion handling: Occlusion is indicated in $F_{k}$ when $N_{k-1} > N_{k}$. The last visible point of a broken trajectory is found in $F_{k-1}$ as the point that fails to link, by extrapolation, to a point in $F_{k}$. Then, the occluded point is generated in the predicted position keeping the number of points constant.