为进一步提高自适应光学系统的成像质量，本文针对目前广泛使用的盲解卷积，相位差法和斑点重建技术开展了深入研究；详细分析了以上三种技术的各自特点、应用场景和处理对象，并结合自适应光学系统的特点，有针对性的加以算法改进；实验采用自适应光学人眼视网膜细胞图像和自适应光学太阳黑子图像进行算法验证，结果表明经改进后的图像处理技术可以有效提高自适应光学图像的质量和分辨力，较好的满足了自适应光学系统对图像事后处理的需求。

In order to further improve the imaging quality of the adaptive optical system, the widely used image reconstruction technologies at present, including blind deconvolution, phase diversity and speckle imaging, are extensively researched in this paper. The characteristics, application scenes and processing objects of each technology are analyzed in detail and the algorithms are improved reasonably combined with the characteristics of the adaptive optical imaging system. Experimental results of in vivo human retinal cell image and sunspots image, proved that the improved technologies can effectively increase the quality and resolution of adaptive optical images, and satisfy the needs of adaptive optics system for image post-processing.