I plan to use either a raspberry pi or a STAR64 (uses risc-v) for AI, most likely will be using a computer that has risc-v in it, the idea is to have a camera hooked up to the mini computer and in real time it can pull all text from the video feed and even possible detect patterns.
I want to use the most optimised library for these two devices as well as ensuring it is Rust.
I know there is opencv rust bindings but is it suitable to use on a risc-v architecture?
And additionally I want to know a good library to use with the computer vision library where I can in real time process what it sees into text, what would be a good library, especially for risc-v architecture?
And lastly I would like to know a good library to detect a similar pattern on objects, lets say there is a design pattern and I want to see with my mini computer if it detects it or not what would be best to use and optimised for risc-v architecture?