Due to the complexity and randomness of image degradation caused by the aero-optics effects，it is difficult for a single restoration algorithm to reach the requirement of restoration. A scheme to construct a compound algorithm to improve the quality of restoration image is proposed, in which some useful algorithms based on different theoretical models and targets are synthesized, and their advantages are combined to make full use of their superiority. Major compound algorithms are given in this paper. The adaptive adjustment and reformation of various algorithm structures as well as the adaptability selection of various parameters, after the combination of statistics theory based MAP restoration algorithms with regularization theory based total variation algorithm, are discussed. The intelligent combination of various algorithm structures is accomplished to achieve their superiority. Some comparisons on restoration were made on microcomputers, and the experiment results show that the proposed compound algorithm improves the overall quality of restored images.