Abstract: Compressive sensing theory is able to solve the problem of data store and transportation introduced by a large bandwidth and an increasing number of channels. Compressive sensing theory is applied to high resolution radar imaging. An compressive radar imaging method based on Regularized Orthogonal Matching Pursuit (ROMP) is proposed and compares with the compressive radar imaging method based on Fast Smoothed L0 (SL0) and L1 norm. Through numerical simulation, the focusing performance of the three algorithms is good and the compressive radar imaging method based on ROMP is much faster than the algorithms based on SL0 and L1 norm.