----
> [!definition] Definition. ([[sheaf image]])
> Let $\mathcal{F}$ and $\mathcal{G}$ be [[sheaf|sheaves]] over a [[topological space]] $X$, and let $f:\mathcal{F} \to \mathcal{G}$ be a [[morphism of (pre)sheaves|morphism of sheaves]]. The **sheaf image** of $f$ is the [[subpresheaf|subsheaf]] of $\mathcal{G}$ given by [[sheafification]] of its [[presheaf image]]. The notation $\im f$ is (still) used.
^definition
> [!basicproperties]
> - Taking images and [[(pre)sheaf stalk|stalks]] are compatible: $(\im f)_{p}=\im(f_{p}:\mathcal{F}_{p} \to \mathcal{G}_{p})$.[^1]
> - If $f$ is an [[injective sheaf morphism]], then the [[presheaf image]] is a [[sheaf]] and [[sheafification|sheafifying]] therefore adds no new information.
^properties
[^1]: It may seem like there is something tricky going on here, since the definition of $\im f$ involves a [[sheafification]]. But recall that sheafification preserves stalks, and this is a result about stalks. So the sheafification does not affect the discussion.
> [!justification]+
> We need to justify that the sheaf image can be naturally identified with a subsheaf of $\mathcal{G}$ — that 'the sheafification can be done inside $\mathcal{G}