The current research is dedicated to line-of-sight rate estimation for a pitch-yaw gimballed imaging infrared seeker mechanism. The line of sight rate estimation problem has a highly nonlinear dynamic that is affected by various error sources such as measurement noise, significant coupling of vehicle channels, and time-delayed seeker detector measurements along with unknown target maneuvers. In order to consider these error sources, a new line-of-sight rate estimation approach is proposed via a two-step strategy. In the first step, an adaptive discrete-time super-twisting observer with an arbitrary convergence time approach is proposed that estimates line of sight rate, in presence of all errors except the delay. In the second step, a delay compensator is presented based on Super-Twisting Observer method to compensate the seeker delay. To increase the estimation accuracy, nonlinear dynamic equations are used and the simulations are implemented to a Software-In the-Loop test bed. The superiority of proposed approach is approved via various simulations and comparison with previously presented methods.