Abstract: Aiming at the problems of large channel estimation calculation overhead and accuracy degradation caused by the maximum delay, Doppler spread, and increase in the number of antennas in the multiple-input multiple-output orthogonal time frequency space (MIMO-OTFS) system, a sensing aided generalized orthogonal matching pursuit algorithm based on atomic threshold (SA-TGOMP) channel estimation algorithm is proposed in this paper. This algorithm first converts the user and surrounding environment information detected by radar into the initial index set of the OTFS channel, and then introduces an iterative strategy of selecting relevant atoms with fixed values and an atomic threshold mechanism to update the support set. Experimental results show that the algorithm proposed in this paper can effectively improve the channel estimation accuracy while reducing pilot overhead.