基于Gold矩阵的计算关联成像研究

A Study of Computational Correlation Imaging Based on Gold Matrix

测量矩阵是决定计算关联成像图像重构质量和成像速度的重要因素之一。本文提出一种新的测量矩阵优化构造方法,利用伪随机序列生成Gold矩阵,该矩阵同时具有Random矩阵的随机性和Hadamard矩阵的确定性。数值模拟和实验结果表明:在满采样条件下,Gold矩阵和Hadamard矩阵均能够实现完美的重构图像;在欠采样条件下,Gold矩阵重构图像质量明显优于Random矩阵和Hadamard矩阵。

Measuring matrix is one of the important factors that determines restructuring quality and speed of computational correlation imaging.In the paper below,a new measuring matrix optimized constructing method was proposed,wherein a Gold matrix with randomness of Random matrix and certainty of hadamard matrix was generated from pseudo random sequences.The result of numerical simulation and experiments shows that both Gold and Hamamard can restructure image satisfactorily under full sampling conditions,but the restructured image quality of God matrix is remarkably superior to those of Random and Hadamard under undersampling conditions.