Micro-lens array type deep learning three-dimensional ghost imaging method based on phase modulation

Abstract

The invention discloses a micro-lens array type deep learning three-dimensional ghost imaging method based on phase modulation. The method comprises the following steps:a neural network model is trained by using a sample set in advance; the broadband thermal light emitted by the target object enters the front imaging module, and forms a two-dimensional image of the target object on the imaging surface of the front imaging module; a micro-lens array type spatial light phase modulation module is arranged at the position of an imaging surface, and a two-dimensional image of a target object is modulated into a speckle field array through the micro-lens array type spatial light phase modulation module; the detection module collects the speckle field array and sends the speckle field array to the image reconstruction module; and the image reconstruction module performs image reconstruction by using a pre-trained neural network model according to the received speckle field array to form a three-dimensional image of the target object. According to the method, the neural network and deep learning are adopted to realize image reconstruction, the calibration process of traditional correlated imaging based on phase modulation is omitted, the reconstruction time can be greatly shortened, and three-dimensional imaging is realized.

Description

technical field [0001] The invention relates to the technical field of quantum correlation imaging, in particular to a microlens array-based deep learning three-dimensional ghost imaging method based on phase modulation. Background technique [0002] With the development of quantum imaging technology, the phase modulation-based passive thermal light source ghost imaging camera is getting closer and closer to commercial applications, but at present, the equipment is bulky, and a complicated calibration process is required before the camera is used. The construction time is long, the use is inconvenient, and it is not conducive to industrialization, especially the three-dimensional correlation imaging based on phase modulation has not been reported yet. Contents of the invention [0003] The purpose of the present invention is: in view of the current passive thermal light source ghost imaging camera based on phase modulation requires a complicated calibration process before ...

AI Technical Summary

Problems solved by technology

[0003] The purpose of the present invention is: in view of the current passive thermal light source ghost imaging camera based on phase modulation requires a complicated calibration process before use, long reconstruction time, inconvenient use, and unfavorable to industrialization, and proposes a phase-based The modulated microlens array deep learning three-dimensional ghost imaging method combines the bionic compound eye and microlens array technology with the phase-modulated passive thermal light source ghost imaging system, and uses artificial neural network and deep learning to realize image reconstruction, which can save the traditional The calibration process of correlation imaging based on phase modulation greatly shortens the reconstruction time and realizes three-dimensional imaging

Images