Adds a concept `CameraEngineConcept` to ensure that camera
engine implementations provide the necessary functions
with the correct signatures and `noexcept` specifications.
This enables compile-time checks for valid camera engine
implementations, improving code reliability and preventing
runtime errors.
Removes the virtual destructor from the Camera class as it is not required,
as the class does not act as a base class. This simplifies the class
definition and avoids potential vtable overhead.
Moves camera and prediction engine implementations into traits for each engine,
decoupling the engine-specific logic from the core classes, promoting code reuse
and maintainability. This change allows for easier addition of new engines and
customization of existing ones.
