Experimental Study On Correspondence Imaging

Ghost imaging is a new imaging technique, it use two beams light coming from one light source, one beam passes through the object and collected by a bucket detector(or point detector), the another beam collected by reference detector after free propagation, the two detectors can reconstruct the image by coincident counts, we can call it correlation imaging as well. Ghost imaging arouses great interest of researchers since it was found. This imaging technique has many advantages than classic imaging. First, it can imaging without lens; Second, it applies to any wavelength; Third, it won’t affect by turbulence and air quality; These advantages and features make it has great potentialities in the use of remote control and military defense. However, ghost imaging also has its own shortcomings. First, thermal light ghost imaging has low visibility, 31 is the maximum for second order thermal light ghost imaging; Second, the image also has a low signal to noise ratio(SNR); Third, the real-time imaging is impossible for now. These limit the practical application of this technology, so researchers study for overcoming shortcomings. A new approach to deal with ghost imaging experiment data were presented in 2012, that is correspondence imaging.According to the way correspondence imaging deal with data, the author did second-order HBT and second-order double-slit interference experiments with pseudo-thermal, there is a point detector on the test arm both in the two experiments, it’s different from the literature using bucket detector. The author uses two parts correspondence imaging to deal with the sampling and reconstruct the image. Further, the author kept the same light source and did another imaging experiment with a bucket detector on test arm. Using four parts correspondence imaging to deal with the sampling from above three experiments respectively, by comparing the reconstructed images can conclude: no matter which kind of detector, the intensity fluctuation has a huge contribution to the imaging. Last, take the two experiments with point detector as example and draw a trend figure reflecting the relationship of visibility and sampling number. By analysis the figure we know that a reasonable threshold can reduce the time to reconstruct image; By comparing the visibility obtained by correspondence imaging with the max visibility( 31) obtained by traditional correlation function method we can find that correspondence imaging make the visibility greatly improved, and further speculate when there is bucket detector on test arm the visibility will also better than traditional correlation function method.