基于离散余弦生成散斑的鬼成像研究

Research on Ghost Imaging Based on Discrete Cosine Generated Speckle

计算鬼成像在进行图像重构时比鬼成像所需要的重构时间短、采样率低,特别是有序投影图案的计算鬼成像能明显地减少采样次数。为进一步研究低采样率下恢复待测物体信息,采用离散余弦变换矩阵作为基底以此来生成投影图案,并采用压缩感知中的TVAL3算法与离散余弦鬼成像相结合,进行数值模拟与实验验证,最后选用图像评价指标对该结果进行评测,最终结果表明:压缩感知中的TVAL3算法与离散余弦鬼成像相结合相比于传统计算鬼成像,实现了在降低采样率的前提下提高成像质量。

Computational ghost imaging requires shorter reconstruction time and lower sampling rate than ghost imaging in image reconstruction,especially computational ghost imaging with ordered projection patterns can obviously reduce the number of sampling times.In order to further study the recovery of the information of the object to be measured at low sampling rate,the discrete cosine transform matrix is used as the base to generate the projection pattern,and the TVAL3 algorithm in compressed sensing is combined with the discrete cosine ghost imaging for numerical simulation and experimental verification.Finally,the image evaluation index is selected to evaluate the result,The final results show that the combination of TVAL3 algorithm and discrete cosine ghost imaging can improve the imaging quality without reducing the sampling rate.