Research Of Quantum Imaging Method Based On Compressed Sensing Theory

As the separation of object and image,ghost imaging is significantly different from classical optics.Due to thermal light is common in nature and easy to obtain, thermal light ghost imaging has become the focus and direction of frontier directions of research in this field in recent years. The principle of remote sensing technology is the classical optical theory. The progress of quantum imaging technology will bring new development of remote sensing physical with its clear images playing an important role in aerospace exploration, military reconnaissance and far-infrared imaging.Based on studying and summing up of lensless ghost imaging with thermal light, we carry out experiments by controlling variable method to research the impact factors of thermal light lensless longitudinal ghost imaging and the relationship between the two channels when the image is clear,and then complete theoretical derivation.We introduce theory of compressed sensing in order to solve the problem of low quality,multiple sampling and long time consuming in traditional lensless ghost imaging with thermal light.We propose a Random-OMP algorithm which is an improved algorithm of OMP algorithm to improve the imaging quality and reduce the computational cost. Obtained results are as follows:1)According to the experimental and theoretical derivation, the transverse size of the thermal light source and the optical path influence the longitudinal coherence scale. With the increase of the transverse size of the thermal light source, the longitudinal coherence scale decreases while the resolution improves and the visibility decreases significantly. Only when the effective diffraction length of the reference channel and the test channel are equal,we can get clear ghost image in the system of lensless ghost imaging with thermal light.2) Based on compressed sensing theory of ghost imaging with thermal light,we use OMP algorithm to get image. Compared with the traditional thermal light ghost imaging, OMP algorithm significantly improvs the quality of image and takes less time without a large number of sample.But the image is still not continuous. When we use improved algorithm Random-OMP to reconstruct ghost image,compared with the OMP algorithm, signal-to-noise and peak signal to noise ratio ratio of the new image increase and mean square value error decreases. The quality of image improvs again with better continuity and lower computational cost. Random-OMP algorithm is more effective in the processing of reconstruction of ghost imaging with thermal light.