Suppose I have two undirected graphs, each with 3 vertices and 3 edges: has = and = , and has = , and = . If you draw these graphs, it's pretty clear that they are equivalent: you get by changing the vertex labels on to . But they aren't equal, since their vertex sets and edge sets are different (in a superficial way).
We capture the idea that and are equivalent except for their labeling by saying they are isomorphic. Two graphs and are isomorphic if there is a bijection such that = , and . What all the symbols mean is that if you can find a way of matching each vertex of with a corresponding vertex of (this matching is your bijection ), and if your matching means that the edges of are mapped to the edges of , then and are isomorphic. Another way of saying the same thing is that and are neighbours in if and only if and are neighbours.
We consider graphs up to isomorphism, that is we consider isomorphic graphs to be equivalent, and a property that we prove about one is true of the other. So we talk about the complete graph of order 5, . Even though there are many isomorphisms, they are all equivalent.