How to implement custom behavior on object detection [closed]

edit

I am currently trying to implement a GPS based navigation. I will be using robot_localization to create the map -> odom -> base_link frame. I have an idea as to how to do the navigation but I was wondering how would I incorporate lanes into it. Lanes will likely not be detected by the Lidar as an object. Is it possible to use a camera to detect lanes and switch behaviours? I would prefer to handle object/lane avoidance behavior on my end. Any reference to documentation/implementation would be helpful here

Edit: It's done using controller plugins in Nav2