A popular imaging technique called Time of Flight (ToF) camera provides a depth information in real time. Several applications like augmented reality (AR) applications, machine automation in factories, hand gesture controls, in military for autonomous navigation and object localization in robotics, ego motion estimation etc. are good source of ToF depth camera. These applications demand good accuracy to provide required services. ToF fulfill this demand of accuracy, however, faces problem like multipath interference (MPI). The MPI phenomena hampers the accuracy of depth map recovery and it can be up to several centimeters. In this paper, we solved the MPI problem by exploiting the sparsity of the received signal. We proposed an approach of sparse regularization technique based on compressed sensing framework with some modification such as applying positive and proximity constraint. This sparse recovery algorithm applied is robust for the MPI with two path. We demonstrate and validate the simulation results of our proposed algorithm for MPI removal.