hi.

im starting a project of AR(augmented reality). there are lots of API and toolkits like wikitude, vuforia and.... problem is most of them has limitations on free version. i would like to understand algorithms and systems behind AR. i want to know is it hard to make a polished product without these toolkits and there are alot behind these toolkits?

I found a blog post that describes the basic technologies behind AR based on what you already mentioned Vuforia and ARKit. YOu should start there but as I read it, it sounds not more difficult as regulary image recognition is so maybe a charge to the AI front may help here