Study on ghost imaging via compressive sensing for a reflected object

Abstract The computational ghost imaging for a reflected object was realized by a spatial light modulator and a coaxial imaging system. The resolution of the reconstructed imaging was improved by the compressive sampling algorithm, and the noise caused by the limited aperture of the lens was minimized by the fuzzy-removing algorithm. After the theory analysis and simulation, the experiment system was set up to verify validity of the algorithm. From the experiment, we can conclude that the reconstructed image of reflected object by compression sensing correlation calculation became clearer with the increase of calculation times. The image obtained by fuzzy-removing algorithm was much clearer than that obtained by none fuzzy-removing algorithm with the same measurement times. Because the noise introduced by the aperture of lens decreased as the increase of the diameter of the lens, the visibility of the reconstructed image increased. The resolution of reconstructed imaging can reach several tens micron order by the compressive sampling and fuzzy-removing algorithm. This method expanded the application of the compressive ghost imaging in the remote sensing, and decreased the complexity of the imaging system in the space platform.