Steps for detecting 3 road signs

edit

This is a rather general problem, and you haven't provided any specifics.

So, in response, here are three answers with no specifics.

1. Define the signs as parts, look for the parts, if you find enough, it's that sign.
2. Define criteria all signs meet, look for that, then look for parts that narrow the category from sign to L, R, or S.
3. Gather lots of sample images, label them, train a Convolutional Neural Network, and run it over your images.

I suggest you start by looking through and trying to understand the various tutorials. It will take a while, but you'll get a much better understanding of both what the problem is, and how it might be solved. Then you can try some of them, and come back with better questions.

You need to provide some more specifics. Does this have to be done in real time? How far in advance of the sign do you need it, etc. Even more important is what sign system are you trying to recognize, Is it US, European, Middle Eastern? Is left hand drive or right hand drive? As far as algorithms, etc, try searching. I found too many good papers to make a suggestion using: machine vision road sign recognition stop sign

I want to detect stop sign red octogon with stop inside, mandatory left a blue circle with white straight arrow inside orientated left, and mandatory right a blue circle with curved white arrow orientated right also I will put camera on robot and it need to recognise the sign at least 1 m before

We can't solve this for you. There are still far too many unknowns. You have to try to solve this on your own. Get to the point where you can say, here is what I am doing, and this is the situation it fails in, can you help me with that situation? That we can do.

Try something simple to start, like finding red octagons and blue circles. If you can find those while being able to differentiate between those and red circles and blue squares (false targets), you'll be almost half done.

There are many, many different ways of doing it, so find one that works for you. Then, once you've done that, it won't work perfectly, it never does. So then come back and ask how to improve it.