Let $$f: (\Omega_1,\mathscr{F_1})\rightarrow(\Omega_2,\mathscr{F_2})$$ be a measurable function. Then we have $f^{-1}(A)\in \mathscr{F_1}$, when $A\in \mathscr{F_2}$.
I am wondering whether the following is true: if $B\in \mathscr{F_1}$ then there exist $A\in \mathscr{F_2}$ such that $B= f^{-1}(A)$.
Not necessarily.
Simple counterexample: $\Omega_1=\Omega=\Omega_2$ with $\mathcal F_1=\wp(\Omega)$ and $\mathcal F_2=\{\varnothing,\Omega\}$ and $\Omega$ is not a singleton.
Let $f$ be the identity function.
Actually it can be shown that the collection $\{f^{-1}(A)\mid A\in\mathcal F_2\}$ is a $\sigma$-algebra. If $f$ is measurable then it is a subcollection of $\mathcal F_1$ but does not necessarily coincide with $\mathcal F_1$ (as shown).