Imaging And Tracking Technology Through Scattering Medium Based On Speckle Correlation

Optical imaging technology is an important way of recording information.It plays an extremely important role in human development.There are various scattering media in life,such as haze,biological tissue,and turbid liquid,etc.Due to the non-uniformity of the refractive index distribution in the scattering medium,the light is scattered when passing through the scattering medium.The interference between the scattered light forms a series of speckles behind the scattering medium.People cannot directly observe the original information of objects from these specklesResearch on how to obtain object information through scattering media becomes particularly important.The object information mainly includes the shape information and the motion state.This corresponds to the imaging and tracking technologies through the scattering medium,respectively.In recent years,wavefront shaping,transmission matrix,machine learning,ghost imaging,and other methods are commonly used to solve the problem,that imaging through scattering media.Among them,wavefront shaping and transmission matrix need to measure the scattering medium in advance.Machine technology needs to train the network with prior input and output.Ghost imaging technology requires the distribution of the illumination light.These have limited their use in practice.Laser illuminated detection and ranging(LIDAR)technology is often used to track moving objects behind corner scattering media.Its high cost and susceptibility to the perturbed environment,however,restrict its applications.The angular memory effect is an optical effect for speckle that was proposed in 1988.This effect indicates that when the incident light shining within the angular memory effect range,the speckles have a strong correlation with each other.In recent years,researchers have demonstrated that this technology can be used to achieve non-invasive imaging of objects through scattering media.Compared with the other techniques mentioned above,the advantage of the imaging technique via the angular memory effect is that it is non-invasive.This imaging technology has great value in real-world applications.Further,the researchers proposed that the angular memory effect can be used to track the movement of hidden objects.Compared with tracking via LIDAR,this method does not require expensive equipment,and its algorithm is more concise.However,there are still some problems with this technique:the efficiency of the image restoration algorithm is low,the role of the speckle bit depth is unclear,the size of the tracked object is limited by the memory effect range,and the accuracy of the tracking algorithm is low.To solve these problems,this article did the following research:(1)The original algorithm has been improved on the speed of image restoration.We studied the original memory effect imaging technology and proposed to use the XX algorithm to restore the object image based on the original algorithm.In this algorithm,there is no need to use "Hybrid Input-Output(HIO)" and "Error reduction(ER)"algorithms,and there is no need to adjust the convergence factor,?,in real-time.This makes the recovery algorithm more concise.At the same time,the new algorithm also greatly reduces the number of iterations and improving the recovery speed.The increase in speed makes the technology have more potential for real-time imaging of dynamic objects.It is a great help to improve the practicality of this technology(2)The effect of speckle depth on imaging and tracking results.In real-time imaging of dynamically hidden objects,a large number of speckles need to be collected.It will take up a lot of storage space.If the object image can be recovered using speckle with a low bit depth,the storage can be reduced.At the same time,the reduction of the speckle bit depth also means that the requirements for the camera hardware are reduced,which has a positive significance for the practicality of the technology.In this article,we studied the relationship between the speckle bit depth and the restoration quality.It is pointed out that even speckles with low bit depth(binary speckles)can be used to restore the object image.And the effect of speckle bit depth on the tracking technology is studied.Reducing the bit depth of speckle will not have a significant impact on the tracking experiments.(3)We revealed a new optical effect of speckles:the "edge memory effect".It has possible to track large objects.When the moving object meets certain conditions,the corresponding speckles still have correlations.We summarized this phenomenon and named it as the edge memory effect.The edge memory effect indicates that even if the size of the object exceeds the memory effect range,as long as the moving distance is within this range,and the overlapped image was taken prior to and after the movement is much smaller than the object itself,the corresponding speckles are correlated to each other.By establishing a model,we explained the edge memory effects.It also demonstrates the existence of edge memory effect through experiments.At the same time,it is shown that the main factor of the speckle cross-correlation tracking of ordinary objects through the scattering medium is that the overlapped image is main interfere with the cross-correlation results.(4)Based on the edge effect,a cross-correlation method of speckle difference is proposed to achieve the tracking of large objects.Because of the conclusion that the overlapped image is the main interference of the cross-correlation results,we propose that the difference method can be used to remove this interference of the overlapped p image.Through analysis,we point out that the difference of the moving objects satisfies the edge memory effect.This makes it possible to calculate the moving distance of the object by cross-correlation.Then,we verified it through experiments.To further improve the practicability of this method,we propose to use threshold filters to filter out inflection points where the objects don't meet the edge memory effect.At the same time,this speckle difference cross-correlation algorithm has an anti-noise ability to environmental noise.(5)Compared with the traditional cross-correlation method,a template matching method is proposed to calculate the speckle movement distance.Through the study of traditional cross-correlation algorithms,we point out that the accuracy of the traditional algorithm is low.The template matching technique can be used to calculate the moving distance of speckles with better accuracy.It is the normalized cross-correlation algorithm.Through numerical simulation,we demonstrate that the new method has higher accuracy in calculating speckle movement.This method is also the method used in the tracking experiment via the edge memory effect.