Most efficient way to create an AI's view?

I'm working on a semi-realistic rts project, and currently focusing on AI. Now, I want the AI to find certain objects within his view, some of them are its enemies of course. Now in the game the so called 'soldiers' are planned to shoot from high places to low places and from low areas to high areas (Literally), have their view blocked by colliders in the map and have a some sort of bound to how far and wide they could see. Now, I bet that there are many techniques to do so, but the most efficient I could think of is to have a child object with a collider that will act as their view bounds, now I imagine that it could be done in a 'better' way, so I ask here for any clues / note-able points that I might use before making this small system.

Many thanks for anyone who'll help out :)