Research On Optical Encryption And Object Classification Based On Ghost Imaging

Quantum imaging,also known as "ghost imaging",is a new way of imaging by using quantum entanglement properties or the intensity fluctuation characteristics of the classical light field.Unlike traditional imaging methods,ghost imaging can recover an image of an object on an optical path that does not contain objects.This property is called non-locality.Therefore,it has become one of the frontiers and hotspots in the field of quantum optics in recent years.we study the application of ghost imaging in optical encryption and object classification.The details are as follows:Firstly,based on the theory of computational ghost imaging and optical encryption,we propose a ghost imaging optical encryption method based on fractional Fourier transform with its unique transform order.In the method,a fractional Fourier transform is introduced into the ghost image such that the designed speckle has the property of a fractional order.When the image of the object is recovered by the received bucket detection value,the image cannot be correctly restored if the predetermined fractional order cannot be accurately known,and therefore the fractional Fourier transform has an encryption function.The simulation and experimental results demonstrate that the feasibility of the ghost imaging optical encryption method based on fractional Fourier transform and have high imaging security.Secondly,we apply the ghost imaging technique to object classification,and propose a method of ghost imaging classification based on linear discriminant analysis.Different from the traditional object classification method,this method needs to design the specific speckle,and can classify unknown objects without obtaining the object in advance.At the same time,the theoretical analysis shows that when ghost imaging is performed by this method,the number of speckles to be irradiated is only the number of categories of objects minus 1,which is more efficient than traditional ghost imaging.We simulate the recognition and classification of handwritten digits and design the speckle for ghost imaging classification according to the Mixed National Institute of Standards and Technology data base(MNIST).The simulation results show that for the recognition and classification of ten categories of handwritten digits,we can classify the unknown digital images with 96.91% accuracy using only 9 speckles.