Note that although the potential difference between the plates is same between any two points of the two plates, the charges on the plates are not uniformly distributed - there will be a different charge density on the plates in the area of \(\epsilon_{r1}\) than in the area of \(\epsilon_{r2}\text{.}\) Let \(\sigma_1\) and \(\sigma_2\) be the corresponding surface charge densities in the two regions of the plates. The charge conservation tells us that
\begin{equation*}
\sigma_1 \:\dfrac{A}{2} + \sigma_2\: \dfrac{A}{2} = \sigma\:A,\ \ \Longrightarrow\ \ \sigma_1 + \sigma_2 = 2\sigma.\ \ \ \ (1)
\end{equation*}
The magnitudes of the electric field between the plates in the two regions occupied by the two different dielectrics will be
\begin{equation*}
E_1 = \dfrac{\sigma_1}{\epsilon_{r1}\:\epsilon_0},\ \ E_2 = \dfrac{\sigma_2}{\epsilon_{r2}\:\epsilon_0}.
\end{equation*}
Now, when we calculate the potential difference between the plates using these two different expressions for the electric field we should get the same value. \begin{align*} \Delta V = \dfrac{\sigma_1\: d}{\epsilon_{r1}\:\epsilon_0}\ \ (\textrm{using } E_1),\\ \Delta V = \dfrac{\sigma_2\: d}{\epsilon_{r2}\:\epsilon_0}\ \ (\textrm{using } E_2) \end{align*} Equating the two we get
\begin{equation*}
\dfrac{\sigma_1\: d}{\epsilon_{r1}\:\epsilon_0} = \dfrac{\sigma_2\: d}{\epsilon_{r2}\:\epsilon_0},\ \ \Longrightarrow\ \ \sigma_2 = \dfrac{\epsilon_{r2}}{\epsilon_{r1}}\: \sigma_1.\ \ \ (2)
\end{equation*}
Putting (2) into (1) helps us solve for \(\sigma_1\) with the result
\begin{equation*}
\sigma_1 = \dfrac{2\epsilon_{r1}}{\epsilon_{r1} + \epsilon_{r2}}\: \sigma.\ \ \ \ (3)
\end{equation*}
Put (3) into either of the two expressions for \(\Delta V\) to get
\begin{equation*}
\Delta V = \dfrac{2\sigma d}{\epsilon_0\: (\epsilon_{r1} + \epsilon_{r2}) } = \dfrac{2Q d}{A\epsilon_0\: (\epsilon_{r1} + \epsilon_{r2}) }.
\end{equation*}
The ratio \(Q/\Delta V\) gives the capacitance \(C\) to be
\begin{equation*}
C = \dfrac{\epsilon_0\: A}{d} \left( \dfrac{\epsilon_{r1} + \epsilon_{r2}}{2} \right).
\end{equation*}