针对激光反射层析成像中目标随机抖动与径向运动造成的投影中心失配问题，提出采用相位恢复算法对重构图像进行相位恢复重建，通过反复光强迭代消除目标随机运动造成的相位误差，达到减少重构误差、恢复目标图像的目的。为改进G-S算法收敛速度与恢复精度，提出频域模值加权方法进行投影数据相位迭代恢复，仿真实验表明，将算法收敛速度与收敛精度提高了1.2倍以上。通过对三组仿真实验处理，重构图像的平均相对均方误差由2.487 5下降到0.792 7，有效地恢复了目标图像轮廓。外场实验表明，该算法能够有效消除重构图像伪迹，改善激光反射层析成像系统成像质量。

The random moving and jitter of the target rotation center will lead to the projection center mismatch in the process of laser reflection tomography. The phase retrieval algorithm was employed to reconstruct the fault image. The phase errors caused by random motion by repeated intensity iteration was eliminated to reduce reconstruction error and restore target image. Aiming at the slow convergence speed and poor accuracy of G-S algorithm，the modulus weighting algorithm was provided to restore projection data. The simulation experiments show that the convergence speed and convergence accuracy of the algorithm are increased by more than 1.2 times. The mean square error of the three sets of reconstructed images in the simulation experiments is reduced from 2.487 5 to 0.792 7. The target image contour also be sharper by the method. The field experiments show that the image reconstruction artifact can be effectively eliminated and the imaging quality of the laser reflective tomography system is improved.