If the world would be in black and white maths was exactly the same that it is.
There is indeed no need of colours in maths at all. The basic fact that justifies this is that we can code maths with only zeros and ones, or in other words, in black and white. In this sense, maths is a science that do not “see in colours”.
In practical terms, if we could create a mathematical machine to simulate a human being it should be unable to see colours, however it could distinguish among them. Let me explain this. It is easy to write a program in order to distinguish colours, and surely it will work well in most of the cases. However, distinguish is different from see. If we take a look to what the machine really sees whenever it watches a colour, we should realise that it is indeed an image in black and white of what a colour is. The reason is that the machine understands in mathematical terms. This makes a huge difference between our perception of the word and the one of a mathematically made machine.
Paradoxically, it is usually said that teaching mathematics needs colours. This is because we are not mathematically programmed machines, moreover, colours allows us to incorporate, understand, remember and differentiate mathematical objects (surely a phycologist could say more about this). More in general, we are used to perceive our surrounding using colours, so, it is not very surprising that representing maths with colours do help us to understand them better (this is an interesting fact, that many maths professors could exemplify vasty).
It is a bit sad that a colour is indeed a not definable object in mathematics, so a proper definition of something so trivial is far beyond what we (mathematicians) can define. Defining a colour is a problem that more successfully than we, physicists, artists, phycologists, physicians and writers have accomplished, however, its fully comprehension looks at least intriguing.