Computational ghost imaging study based on incoherent light from blackbody radiation

Abstract

In recent years, ghost imaging has made important progress in the field of remote sensing imaging. In order to promote the application of solar ghost imaging in this field, this paper studies the computational ghost imaging based on the incoherent light of blackbody radiation. Firstly, according to the intensity probability density function of blackbody radiation, the expression of contrast-to-noise ratio (R_CN) describing the quality of computational ghost imaging is obtained, and then the random speckle pattern simulating blackbody radiation is generated by computer with the idea of slice sampling, finally, a digital light projector is used to modulate and generate the random modulated light that simulates the blackbody radiation light source, and this light source is used to realize the computational ghost image of the reflective object in the experiment. The “ghost image” of the object under different measurement frame numbers is reconstructed, and the contrast-to-noise ratio describing the imaging quality is measured. The results show that the image quality is relatively good when the average intensity (gray) of the randomly modulated speckle is about 160. On the other hand, the contrast-to-noise ratio of the image gradually increases from 0.8795 to 1.241, 1.516, 1.755, 2.100 and 2.371 as the number of measurement frames increases from 2,000 to 4,000, 6,000, 8,000, 12,000 and 20,000, respectively. The experimental results are basically consistent with the theoretical analysis. The results are of great significance for the application of ghost imaging with incoherent light, such as sunlight, which is approximately regarded as blackbody radiation, in the field of remote imaging.