Chen Zhipeng, Shi Jianhong, Zeng Guihua
Appl Opt. 2016 Oct 20;55(30):8644-8650. doi: 10.1364/AO.55.008644.
Ghost imaging is a transverse imaging technique that allows an object to be reconstructed using the correlation between a pair of light fields. As known, in ghost imaging configurations, a large number of realizations are usually required for reconstruction of the objects. To reduce the number of realizations, Chen et al. [Opt. Lett.38, 546-548 (2013)OPLEDP0146-959210.1364/OL.38.000546] demonstrated an object authentication method with computational ghost imaging using realizations of less than 5% of the Nyquist limit. In this paper, we have further improved this method using a "compressive sensing algorithm" instead of a "classical correlation algorithm in computational ghost imaging." As a result, the realizations for object authentication were further reduced from 5% of the Nyquist limit to 3% of the Nyquist limit.
鬼成像技术是一种横向成像技术，它能够利用一对光场之间的相关性来重建物体。众所周知，在鬼成像配置中，通常需要大量的实现来重建物体。为了减少实现的数量，Chen等人[《光学快报》38, 546 - 548 (2013) OPLEDP0146 - 959210.1364/OL.38.000546]展示了一种使用计算鬼成像的物体认证方法，该方法使用的实现次数少于奈奎斯特极限的5%。在本文中，我们使用“压缩感知算法”而非“计算鬼成像中的经典相关算法”进一步改进了该方法。结果，用于物体认证的实现次数从奈奎斯特极限的5%进一步减少到了奈奎斯特极限的3%。
