An efficient solution of the perspective problem via a suitable delay Riccati equation

In this paper it is proposed a new approach to the solution of the classical state estimation problem for a general time-varying perspective system. The solution is achieved by the definition of a virtual output which has the property to convert the original nonlinear output measurement function into a linear time-varying one. This allows the computation of an efficient observer gain matrix by the real-time solution of a suitable delay Riccati equation. Numerical results show high performances of the observer.