Laser reflective tomography imaging is a new approach of long-range and high-resolution imaging. Reflective echo can be approximately considered as the adding noise convolution of transmitted pulse and target reflectance projection distribution. Hence, the broadening of transmitted pulse has significantly effect on resolution of reconstruction image. The existing filtering back-projection algorithm is insufficient to solve this problem effectively, while the pulse compression and deconvolution method can extract target reflectance distribution projection from echo, which will enhance the quality of reconstruction images. Based on the similarity of multi-angle detection, the blind deconvolution of one-dimensional echo signal is proposed based on multi-frame iterative approximation. Experiments are conducted on real laser radar, the reconstructed results show that the proposed method can significantly enhance the resolution and details of imaging, and give reference for the transmitted pulse.