Color calculation ghost imaging method based on deep learning

Abstract

The invention discloses a color calculation ghost imaging method based on deep learning. The method comprises the steps: continuously projecting coded color illumination speckle patterns to an object through a color light source, collecting the reflection or transmission light intensity of the object through a single-pixel detector, and generating an information graph according to an information recovery algorithm; corresponding information graphs are generated for known different objects, so that a training data set is obtained; constructing a generative adversarial network model, and performing comparison training on the information graph and the object image through a deep learning method to obtain trained network parameters; and finally, inputting the information graph of the to-be-imaged object into the trained network to obtain a corrected and improved color image of the to-be-imaged object. Due to the fact that a random coding mode is used in the method, the complexity of illumination speckle coding in color ghost imaging is reduced; meanwhile, according to an information recovery algorithm provided by the invention, color information in an information graph is reconstructed by using a deep learning method, and a good color ghost imaging effect on an object is realized.

Description

technical field [0001] The invention belongs to the technical field of computational ghost imaging, and specifically designs a color computational ghost imaging method based on deep learning. Background technique [0002] Ghost imaging is a new imaging technology that can acquire target image information non-locally by measuring the intensity correlation function between the reference light field and the target detection light field based on the correlation characteristics of light field fluctuations. [0003] Computational ghost imaging is the development and extension of traditional ghost imaging. [0004] The computational ghost imaging theory points out that the purpose of the reference optical path is to measure the light intensity distribution of the light field reaching the imaging object, so the light field can be regulated by a spatial light modulator or a digital micromirror, and then the light field arrival can be obtained by calculation. Intensity distribution w...

AI Technical Summary

Problems solved by technology

[0006] If only a single-pixel detector is used to detect the light intensity of the signal light, the overlapping of colors will cause confusion of different color information during the imaging process, and the image and color of the object cannot be reconstructed correctly.

[0007] Recently, researchers have proposed a computational ghost imaging method based on special color speckle fields. Although this method uses a single-pixel detector to image color objects, it requires special calibration and coding for different color objects.

Images