Image Understanding

Returning to the problem of having the same patterns at x and x + p. As indicated in eq. (2.21), here it has

$E\left[S_t\left(x;t_i\right)\right]=E\left[S_t\left(x;t_i+\frac{p}{B_i\lambda }\right)\right]=2N_W{\sigma }_n^2\left(2.30\right)$

Note that the uncertainty problem still exists, as the minimum occurs at tp = ti + p/(Biλ). However, with the change of Bi, tp changes but ti does not change. This is an important property of SSSD in inverse-distance. By using such a property, it is possible to select different baselines to make the minima appear at different locations. Taking the case of using two baselines B1 and B2(B1 ≠ B2) as an example, it can be derived from eq. (2.29) that

$\begin{array}{l}E\left[S_{t\left(12\right)}^{\left(S\right)}\left(x;\widehat{t}\right)\right]={{\displaystyle \sum _{j\in W}\{f\left(x+j\right)}}^{}\end{array}$