Alternate perspective other than probabilistic perspective to implement slam

Question

I am bother on some line which is written is Probabilistic Robotics book by Dr.Sebestian Thrun From a probabilistic perspective, there are two main forms of the SLAM problem,which are both of equal practical importance. One is known as the online SLAM problem: and offline SLAM problem

I want to know that Is there any other perspective of slam? Because the author write that from a probabilistic perspective. If there any other perspective of slam please let me know.

Thank you

Answer 1

SLAM algorithms are generally application-specific implementations of the extended Kalman filter (EKF), in my opinion. Because EKF relies on probabilistic estimates of model states given noisy data, I don’t know how to understand SLAM without also understanding the probabilistic underpinnings of EKF. You might benefit from studying various presentations of EKF to see if any of those treatments describe the concepts with less reliance on the statistical framework.