基于稀疏约束的鬼成像激光雷达

发布时间：2018-01-28 02:08

本文关键词： 激光雷达经典鬼成像计算鬼成像压缩感知压缩鬼成像　出处：《南京理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：传统的激光雷达在远距离目标探测中很难同时满足成像速度、灵敏度以及抗干扰能力等方面的性能需要。然而,鬼成像理论可以使得激光雷达具备更强的抗干扰性能;压缩感知理论则可以大大减少采样次数,提高成像速度。本文将这两者结合,研究了压缩鬼成像算法,并基于该理论搭建了一套雷达样机系统,探索了它在实际目标探测成像中的应用,具有更加优异的成像性能。主要内容包括:1)经典鬼成像理论的数学推导:本文通过符合测量和二阶关联理论,推导了鬼成像的数学关系式;研究了赝热光源的数学表达,阐释了其具有和真热光源相似的特性,证明了赝热光源鬼成像的可行性。2)压缩鬼成像算法的原理:本文首先从统计学角度,阐述了计算鬼成像的实现原理;接着,解释了压缩感知理论,指出常见探测目标满足压缩感知所需要的稀疏性条件,进而推导了 CS算法在计算鬼成像中应用的可行性,就此提出压缩鬼成像算法;最后,通过二值目标和灰度目标的仿真以及实验,验证了该算法。3)稀疏约束鬼雷达的样机研制:本文基于压缩鬼成像算法搭建了一套激光雷达样机系统,该系统结构精简、原理清晰。通过选取远处靶标进行探测表明,稀疏约束鬼雷达在成像速度和质量等多个方面相对于传统激光雷达具有明显优势。4)鬼成像中的散斑正交化研究:赝热光源中毛玻璃形成的散斑场呈复高斯圆分布,所以对应的散斑强度满足二项式分布,虽然满足算法需要,但其非正交的特性限制了成像质量。因此,提出一种散斑正交化的方法,将二项式分布转换为正交分布。该方法在不增加额外采样时间的条件下,显著地提高了重建目标的信噪比。
[Abstract]:It is difficult for traditional lidar to simultaneously meet the requirements of imaging speed, sensitivity and anti-jamming ability in long-range target detection. Ghost imaging theory can make lidar have stronger anti-jamming performance. Compression sensing theory can greatly reduce the number of samples and improve the imaging speed. In this paper, the compression ghost imaging algorithm is studied, and a radar prototype system is built based on the theory. Its application in the actual target detection imaging is explored, and it has better imaging performance. The main contents include the mathematical derivation of the classical ghost imaging theory: the coincidence measurement and the second-order correlation theory are adopted in this paper. The mathematical relation of ghost imaging is deduced. The mathematical expression of pseudo thermal light source is studied, and its characteristics similar to that of true heat source are explained. The feasibility of ghost imaging with pseudo thermal light source. 2) the principle of compressed ghost imaging algorithm is proved. The principle of calculating ghost imaging is described. Then, the theory of compressed sensing is explained, and the sparsity condition of compressed sensing is pointed out. Then, the feasibility of CS algorithm in calculating ghost imaging is deduced. In this paper, a compressed ghost imaging algorithm is proposed. Finally, through the simulation and experiment of binary target and gray target, it is verified that the algorithm. 3) the prototype of sparse constrained ghost radar is developed. In this paper, a set of laser radar prototype system based on compressed ghost imaging algorithm is built. The structure of the system is simple and the principle is clear. Sparse constrained ghost radar has obvious advantages over conventional lidar in imaging speed and quality. Speckle Orthogonalization in Ghost Imaging: the speckle field formed by the ground glass in the pseudo-thermal light source is distributed in a complex Gao Si circle. Therefore, the corresponding speckle intensity meets the binomial distribution, although it meets the needs of the algorithm, but its non-orthogonal characteristics limit the imaging quality. Therefore, a speckle orthogonalization method is proposed. The binomial distribution is transformed into orthogonal distribution, and the SNR of the reconstructed target is significantly improved without adding extra sampling time.
【学位授予单位】：南京理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN958.98

【参考文献】