As has been said, group homomorphisms are maps between groups which respect structure.
This is easy to picture in the context of isomorphisms between groups. The size of the two groups is the same (even if they are infinite), and their structure behaves the same (even if the underlying sets and binary operations are different). So finding an isomorphism is basically like translating a passage between two languages.
But it's not so easy to picture when the maps aren't bijective.
Sometimes you'll be embedding a group into a bigger group, in other words constructing an injective homomorphism that isn't surjective. That's not too hard to picture though, it just looks like this.
![enter image description here](https://cdn.statically.io/img/i.sstatic.net/QXNsc.png)
You see this is just an isomorphism to a subgroup of the target group. Injective homomorphisms don't need to be unique in this case, for example lots of groups have many subgroups isomorphic to $C_2$, so you could embed $C_2$ in a bunch of different places through different injective homomorphisms.
Surjective homomorphisms (which aren't injective) are a bit of a different story. Here's what they look like.
You might recall the first isomorphism theorem, that if $\varphi:G\rightarrow H$ is a homomorphism, then there is another homomorphism $\theta:G\rightarrow G/\mbox{Ker}\varphi$ and an isomorphism $\mu:G/\mbox{Ker}\varphi\rightarrow \varphi[G]$ so that $\varphi=\mu\theta$. (In the case pictured above, $\varphi$ is surjective, so $\varphi[G]=H$.) This is a complicated looking theorem but it is actually quite well illustrated above. Intuitively, it means you can divide the group up into those colored blocks which behave collectively the same way in $G$ as their images do in $H$. Remember that all homomorphisms can be viewed this way, and that in fact an equivalent formulation of a homomorphism is that $\varphi:G\rightarrow H$ is a homomorphism if and only if $\mbox{Ker}\varphi$ is a normal subgroup of $G$. Even injective homomorphisms can be thought of like this, but by definition the kernel of an injective homomorphism is trivial, so the blocks are just one element. It is important to understand this "contracting" of blocks of elements to understand homomorphisms.
Now, there are homomorphisms that are neither injective nor surjective, but they can be pictured easily by combining the above two diagrams. Colored blocks on the left will correspond to preimages of colored blocks on the right as in the second diagram, but like the first diagram they won't cover the whole target group (i.e. there are still some grey blocks that never get mapped to).
So having covered group homomorphisms in detail, it is important to realize that homomorphisms in other algebraic structures behave in exactly the same way. Module, ring, and field homomorphisms all obey the same isomorphism theorems. Though it is a little harder to visualize with more than one operation (and hence more than one multiplication table), they all behave like this, and obey the same "contraction" principle.