Investigation And Application Of Correlated Imaging In Non-line-of-sight Imaging

With the development of detection methods and optical systems,the detection of target objects is becoming more and more accurate.However,how to get reconstruction well in environments full of noise has been the focus and difficulty of current research.Non-line-of-sight(NLOS)imaging,as a method which can image in a special environment,does not require target objects to be in the visible range.This particularity let it have great potential in many areas,such as search,rescue,counter-terrorism,driverless,biomedical imaging,etc.Correlated imaging,an alternative imaging method,can obtain the information of object by correlating light field,and it has a great application prospect in NLOS imaging.This thesis mainly explores the application of correlated imaging in NLOS,and divide correlated imaging into spatial correlated imaging and temporal correlated imaging,or Ghost imaging(GI)and Time-Correlated Single Photon Counting(TCSPC).Discuss the role of the two technologies in the two scenes of "through-scattering medium imaging" and "imaging behind corner".Spatial correlated imaging,detecting the correlation between the light field and the spatial distribution of targets,no longer needs to meet the imaging relationship.so it has a high anti-noise ability.Not only that,the scattering effect is not the key factor when through scattering medium in GI.Therefore,it has the ability to imaging through scattering medium.This thesis first builds an experimental system of spatial correlated imaging,by using the spatial light modulation device that can realize high-speed binary modulation to control the structured light field,and using a bucket detector with only one pixel to collect the signal.and restoring the object according to the correlated algorithm.Then a comparative experiment with traditional CCD imaging method is conducted to explore the application potential of correlated imaging in imaging through scattering media.Finally,a method based on GI system is proposed to expand the field of view,which uses the scattering effect to offset the limitation of the field diaphragm to a certain extent.Due to its extremely high sensitivity and response capability,temporal correlated imaging can detect very weak signals,which has a great role in "imaging behind corner".The principle comes from the TCSPC technology.The temporal correlation of photon signals is used to obtain the distribution of light signals in time by statistical method.This method has extremely high time resolution and sensitivity,and gets rapid development in the field of lidar.Combined with the back-projection algorithm,it can achieve the information of object outside the sight,and its positioning accuracy can reach the order of centimeters.In the thesis,detecting object out of sight based on temporal correlated imaging is simulated,and we analyze the influencing factors and deficiencies of this system.Subsequently,the experiment is built to detect the target object that hiding in corner,by using ultra-fast laser,single photon detector,and TCSPC.