To solve the problem that the laser echo is easily polluted by noise and the sample will be lost randomly when there is occlusion, which will affect the extraction of useful information in the echo, a method of laser echo signal enhancement based on the bayesian factor analysis (BFA) model is proposed. Firstly, based on the analysis of the characteristics of the laser echo, the orthogonal factor model is used to model it, so that the problem of signal enhancement can be transformed into the problem of solving the model parameters. Then the variable bayes expectation maximization (VBEM) algorithm is used to update the model parameters adaptively and iteratively, and the laser echo enhancement can be realized under the minimum reconstruction error criterion. Finally, the measured data are utilized to carry out the experiment. In the environment of low SNR and incomplete random missing data, the proposed method can obtain better noise suppression performance and higher missing sample reconstruction performance, so as to ensure that the follow-up signal processing can accurately collect high-value information in the echo.