论文部分内容阅读
二维编码阵列是编码孔径成像的关键部件,它直接决定着再现的层析图像的质量。目前仍没有一种理想的二维阵列既具有较高的量子收集率,又具有良好的层析成像特性。采用一种新的方法———分割矩阵(DIRECT)全局优化算法,设计二维阵列,该算法适用于多变量“黑盒”问题的求解,并且具有比其他优化算法更快的收敛速度。其目的是设计一类自相关函数旁瓣最大值为1,同时具有最大填充率的二维编码阵列。理论分析及实验结果表明:用该算法搜索得到的二维阵列既具有较高的量子收集率,又具有良好的层析成像特性。
Two-dimensional code array is a key part of coded aperture imaging, which directly determines the quality of the reproduced tomographic image. At present, there is still no ideal two-dimensional array which not only has high quantum yield but also has good tomographic features. A new method named DIRECT global optimization algorithm is used to design two-dimensional array. The algorithm is suitable for solving the multivariable “black box” problem and has a faster convergence speed than other optimization algorithms. Its purpose is to design a two-dimensional code array with maximum sidelobe of autocorrelation function and maximum filling rate. Theoretical analysis and experimental results show that the two-dimensional array obtained by this algorithm has high quantum collection rate and good tomographic characteristics.