Research On Real-time Ghost Image Reconstruction Based On Multidimensional Walsh Transform

Imaging quality and imaging efficiency have always been the goals of scientific researchers.With the development of imaging technology,ghost imaging gradually enters the public's field of vision due to its unique advantages.Ghost imaging,also called associative imaging,is a new type of optical imaging technology.Ghost imaging can achieve high-resolution imaging under conditions of atmospheric turbulence and extremely low light,and has strong anti-interference ability and low cost.As we all know,the sampling and imaging of traditional film cameras and modern digital cameras can be completed in the same area,while ghost imaging separates the two processes of sampling and reconstructing images,and finally realizes "Space imaging".Two-dimensional Walsh transform is used in this article to complete real-time imaging experiments and simulations of moving objects.First of all,through comparative analysis,it is known that the two-dimensional Walsh transform is simpler than the one-dimensional Walsh transform,the one-dimensional Hadamard transform and the second-order correlation calculation.It has faster imaging speed and stronger energy concentration.Since the two-dimensional Walsh transform ghost imaging scheme cannot adapt to the existing one-dimensional Walsh transform system,the Walsh matrix obtained by the Kronecker product method is used as the illumination light field in this article.In this way,the one-dimensional Walsh transform sampling process can be equivalent to the two-dimensional Walsh orthogonal transform of the target image.The bucket detection value obtained after transformation can be used to obtain a high-quality reconstructed image using the two-dimensional Walsh transform inversion method.Through simulation,it is known that the two-dimensional Walsh transform has better energy concentration,and the high-energy values in the transform coefficient matrix are concentrated in the upper left corner.Therefore,the zigzag scanning method used in this article is used to obtain a certain number of bucket detection values in the bucket detection value matrix to reconstruct the target object.From the separability characteristics of the two-dimensional Walsh transform,it can be seen that the fast two-dimensional Walsh transform algorithm can be obtained by two fast one-dimensional Walsh transform algorithms,thereby further accelerating the speed of reconstructing images.The base pattern selected in this article is a fully positive base pattern(only two values,0 and 1).The base pattern is obtained through (W_N²xN²+1)/2,and the base pattern is also reordered according to the zigzag scanning method.Use the sorted base pattern to illuminate the target object to obtain the corresponding bucket detection value.This article derives that the barrel detection value obtained by the all-positive base pattern can be used to obtain the barrel detection value obtained by irradiating the target with the W_N²×N² base pattern,which is finally used to reconstruct the image.Using this method can reduce the number of base patterns,thereby reducing the sampling time.Through numerical simulation and experimental verification,it can be known that the scheme in this paper has good reconstruction performance for high-resolution target objects.When the number of samples is set to 880,the real-time imaging of moving objects is realized by the method in this paper under the premise of ensuring the imaging speed and quality.The energy concentration and orthogonality of the two-dimensional Walsh transform in this paper are used to complete the task of reconstructing high-quality target images with a small number of bucket detection values,and the inverse reconstruction method and two-dimensional fast Walsh transform are used to improve the imaging performance.speed.Finally,real-time imaging of moving objects is completed.