1: \begin{definition}

2: The fill distance associated with a set $\mathcal X\subset \partial \Omega$ is defined by 

3: \begin{equation}

4: h_{\mathcal X,\partial \Omega} \coloneqq \adjustlimits\sup_{\Vy\in \partial \Omega}\inf_{\Vx\in \mathcal X} d_{\partial \Omega}(\Vx,\Vy),

5: \end{equation}

6: where $d_{\partial \Omega}$ denotes the geodesic distance associated with $\partial \Omega$.

7: Note that $d_{\partial \Omega}$ is equivalent to the Euclidean metric restricted on $\partial \Omega$

8: \cite[Theorem 6]{FUWR12}, that is, there are two constants $c_1,c_2>0$ (that depend only 

9: on $\partial \Omega$) such that

10: \ben\label{eq:distance}

11: c_1 \|\Vx-\Vy\| \le  d_{\partial \Omega}(\Vx,\Vy) \le c_2 \|\Vx-\Vy\|

12: \quad\text{ for all }  \Vx,\Vy\in \partial \Omega\,.

13: \een

14: \end{definition}

15: