What is the most important is to separate the scatterers and keep their own phases simultaneously in the research of three-dimensional (3D) interferometric imaging for space micro motion target, however, when the pulse repetition frequency (PRF) does not satisfy the needs of the Nyquist Sampling Theorem, the image processing based method can't effectively separate the scatterers. Therefore, a 3D interferometric imaging method based on sparse dictionary decomposition is proposed. The method can be used to separate each of the scatterers from echo data directly. First of all, according to the characteristics of target echo signal, an over-complete sparse dictionary is constructed, and then the echo is decomposed by using sparse decomposition algorithm to get the sub-echoes of each scatterer. Secondly, by means of time-frequency analysis and its performance of phase preservation, the micro motion curves of each scatterer are obtained, and the interferometric phase differences in the time-frequency plane are extracted. Finally, according to the relationship between interference phase differences and coordinates, the 3D imaging of space spinning target is constructed. The simulation results show that when PRF is greater than 0.25 times of Nyquist frequency, the proposed method can effectively achieve the 3D imaging of space spinning target.