Research On Correlated Imaging Technology Under Photon Counting Condition

Correlated imaging is a new imaging technique that uses the second-order correlation calculation of the light field to perform target image reconstruction.Due to its non-locality,high resolution,and anti-jamming characteristics,its imaging methods have received great attention since the concept of correlated imaging.When the light intensity of the light source is large,the correlated imaging system can not consider the detection model carefully,and the light intensity of the signal can be collected by a barrel(point)detector with no resolution capability,without considering the detection limit and the influence of external environmental noise.When the light intensity of the light source is weak to a certain extent,especially in biomedical applications,the intensity-correlation imaging scheme used must consider the detection model of the signal light.Also,the influence of ambient noise and detector dark noise cannot be ignored at this time.Based on the photon counting strategy to analyze the performance of the correlated imaging system in the case of a small number of photons,quantitative analysis of the influencing factors and put forward an improved program for the association of photon counting counts in biomedical,remote sensing detection.The applications and developments in many fields have great practical significance.This paper studied in detail the related problems of correlated imaging systems in the case of photon counting and obtained the following main results:Firstly,based on the classical optical theory,the imaging mechanism and influencing factors of the pseudothermal correlated imaging system are studied.The optical transmission characteristics and regularity of the pseudothermal correlated imaging system are obtained.According to the different representations of the barrel detector and the point detector,the classical lensless correlated imagingsystem is theoretically analyzed,and different optical system transfer functions are obtained.When the imaging distance of the dual-optical path image is satisfied,the distance from the reference light detector to the light source is equal to the distance from the object plane to the light source.Whether it is a barrel detector or a point detector,an image of the object to be measured is reconstructed by an association calculation.When meeting the requirement of double optical path Fourier transform image imaging distance,the distance between the reference light detector and the light source is the same as the distance of the signal light detector.The result based on the bucket detection calculation is an enlarged blurred object image;however,the result calculated by the point detector is the Fourier transform image of the object.Next,using the step-by-step Fourier algorithm to simulate the analysis of the optical transmission part of the correlated imaging system,and verify the impact of the size of the light source,transmission distance on the correlated imaging.The research results show that for the barrel detection scheme,under the premise that the two detectors are at the same distance from the light source,as the object moves from the signal light detector to the light source,the object image gradually becomes large and blurred.For the point detector scheme,the object image gradually changes to a Fourier transform image.When increasing the lateral size of the object,the Fourier transform image is more difficult to obtain,which matches the diffraction theory.Secondly,a non-negative least squares correlation imaging method and a limit quantization correlation calculation scheme are proposed.It is found that the higher quality imaging results can be obtained with fewer sample numbers and calculation time.First of all,several typical correlation imaging methods are studied,namely the traditional second-order correlation imaging algorithm,differential correlated imaging algorithm,compressive sensing correlated imaging algorithm,and pseudo-inverse correlation imaging algorithm.Next,aiming at calculating the correlated imaging scheme,the experimental results verify the correctness of the algorithm,analyze the imaging quality and the time consumption of the algorithm.A non-negative Least Squares correlation imaging algorithm was proposed based on the imaging quality and the number of samples and the time consumption of the algorithm.The results show that the non-negative least-squares correlation imaging algorithm performs well in numerical calculations,and the number of samples required relative to compressive sensing and pseudo-inverse correlation imaging algorithms is the smallest,and the imaging quality is also the best.Finally,from the simulation and experimental results,the characteristics of the signal light data obey the Gaussian distribution,and according to the distribution characteristics of the signal light data,the limit quantization correlation imaging scheme is proposed.The correlated imaging calculation method describes the correlation calculation between the reference light and signal light intensity fluctuations.Therefore,the larger fluctuation measurement sample in the signal light data contributes more to the imaging result,and the fluctuation in the signal light data is smaller.The measurement sample basically has no effect on the imaging result,but instead increases the calculation amount.The finite quantized correlated imaging scheme describes that the signal light data is quantized,and then the quantized data is correlated with the reference light,and the high-quality image of the target object can still be reconstructed.In addition,the limit quantized correlation imaging was compared with the corresponding imaging scheme.The proposed quantized correlation imaging scheme proposed in this paper can be used to store or transmit compressed signal optical data.Thirdly,the correlated imaging system model was established under the photon counting condition.The influence of detector noise on the correlated imaging under the condition of photon counting was studied.Both simulation and experimental results showed that the correlation image quality can be improved by increasing the ratio of signal light to noise.Based on the application potential of the computational correlation imaging scheme at the present stage,only the imaging mechanism when the signal light intensity is low is analyzed.When the signal light intensity is weak,the signal light detector is no longer suitable for using a CCD as a barrel detector.Therefore,a photon counting module with a higher sensitivity is used as a collection device for signal light.When the signal light intensity is very weak,the introduction of randomly distributed detector noise floor will increase the background of the imaging results,and even the target image cannot be obtained.The research shows that the ratio of the signal light average intensity to the detector average noise intensity has an important influence on the correlated imaging results.With the increase of the average signal intensity and the average noise intensity of the detector,the signal-to-noise ratio and the visibility of the imaging results are correspondingly improved.Moreover,the experimental results are well matched with theoretical analysis and numerical simulation results.Fourthly,due to the detector noise will reduce the quality of correlated imaging under photon counting conditions,two noise suppression schemes are proposed: Based on the distribution characteristics of signal light data,a light intensity fluctuation is selected in correlated imaging scheme.Studies have shown that this method can reduce calculations.Time and improve the imaging efficiency;Based on the distribution characteristics of noise,an improved compression sensing algorithm is proposed.It is found that this method can suppress the influence of noise on the correlated imaging.Firstly,based on the Gaussian distribution characteristics of signal light data,in the photon counting correlation imaging system,a light intensity fluctuation selection correlation imaging scheme is proposed,and the number of measurement samples is reduced without reducing the imaging quality.Secondly,based on the characteristics of random detector noise data obeying the Gaussian distribution,an improved compression-sensing correlation imaging algorithm is proposed to suppress and eliminate the effect of noise on imaging quality.Experiments have shown that correlated imaging under photon counting conditions can be negatively impacted by detector noise.Therefore,suppressing noise and improving imaging quality are the primary tasks of the photon counting correlation imaging scheme.By appropriately selecting the thresholds in the algorithm,the effects of noise can be greatly reduced and the image quality can be improved.In addition,this method is also applicable to the imaging of different sparsity target objects and grayscale objects.