Research On Fast Focusing Method For Multi-depth Targets In Computational Ghost Imaging

Ghost imaging(GI),a new method of imaging using the correlation of light intensity rise and fall,has certain advantages over traditional imaging methods and is one of the hot topics of current research.Computational ghost imaging(CGI)is an improvement of traditional ghost imaging,which can obtain the axial depth of a target by evaluating the blurring degree of the reconstructed image.However,the method is still plagued by many problems,such as background noise that can interfere with the stability of the evaluation function's operation,a high number of samples and a high number of reconstructed images,all of which limit the practicality of computational ghost imaging depth estimation methods.In addition,the current computational ghost imaging system has not yet achieved the estimation of multi-depth targets.The problem is that the reconstructed images of multi-depth targets may observe the focus/outof-focus phenomenon of different targets at the same time,thus affecting the performance of the evaluation function.To address these problems,the text proposes a fast focusing method for multi-depth targets based on a computational ghost imaging system,with the main work as follows.(1)Based on the statistical and spatially coherent properties of the pseudo-thermal light source,the pseudo-thermal light source ghost imaging system and the reconstruction method of the computational ghost imaging system are investigated,and the propagation properties of the scattered spots in the deep Fresnel region and the focus/off-focus model of the computational ghost imaging in the computational ghost imaging system are analysed.(2)A target depth estimation method based on an adaptive focusing window is proposed.A search interval is first selected based on the overall characteristics of the evaluation function,and then an iterative search is performed to find the target axial depth within the selected interval.During the iterative process,the adaptive window is designed to ensure the integrity of the target within the window while effectively reducing the background region.Experimental results show that the method significantly reduces the necessary working distance required for the evaluation function,making it equally applicable under undersampling conditions,and also reduces the influence of background noise on the evaluation function,enhancing the robustness of the algorithm and further improving the depth estimation method based on computational ghost imaging systems.(3)The characteristics of out-of-focus images of multi-target scenes in the spatial domain are analysed,the gradient domain is selected as the image transformation space,and the image reconstruction is reconstructed by a compressing sensior algorithm based on the TV model,which suppresses the interference of out-of-focus targets in the image,so as to estimate the depth at which each target is located.In addition,using the correlation between the scattered horizontal coherence distance and the vertical coherence distance,a variable resolution depth estimation method is proposed,which reduces the required depth images by approximately 55%within a given system,effectively improving the efficiency of the algorithm.Based on the computational ghost imaging system,this paper firstly achieves the enhancement of the depth estimation algorithm for 2D targets,and then proposes the first fast focusing method for multi-depth targets,which strengthen the knowledge and understanding of the nature of ghost imaging systems,advances the development of ghost imaging autofocus methods,and helps ghost imaging technology to move faster towards practical applications.