论文部分内容阅读
激光反射断层成像是一种新型的远距离高分辨率成像手段。反射回波可以近似视作发射脉冲和目标反射系数投影分布的加噪卷积,因此发射脉冲的展宽将显著影响重建图像的分辨率。现有的滤波反投影方法不能有效解决这个问题,而脉冲压缩和解卷积可以从回波中还原目标反射率分布投影,进而提高重建图像质量。基于多角度测量的相似性,提出了一种多帧迭代盲解卷积逼近方法,对一维回波信号进行解卷积,并在实际激光雷达上进行了实验,重建结果表明,此算法能显著提高成像的分辨率和细节,并能对发射脉冲给出参考。
Laser reflection tomography is a new type of long-range high-resolution imaging method. Reflected echoes can be approximated by the noise-added convolution of the transmitted pulse and the projected reflection of the target reflection coefficient, so the broadening of the transmitted pulse will significantly affect the resolution of the reconstructed image. The existing filtering backprojection method can not effectively solve this problem, and pulse compression and deconvolution can reduce the projection of the target reflectivity distribution from the echoes, thereby improving the quality of the reconstructed image. Based on the similarity of multi-angle measurement, a multi-frame iterative blind deconvolution approximation method is proposed. One-dimensional echo signals are deconvoluted and experimentally performed on an actual laser radar. The reconstruction results show that this algorithm can Significantly improve imaging resolution and detail, and can give a reference pulse launch.