Wave Field Parameter Inversion Algorithm Based On Radon Transform And Image Reconstruction

CT(Computer Tomography)technology has been used more and more since its introduction.Therefore,computer tomography is a research direction.The current goal of CT development is to continuously improve the speed and quality of image reconstruction.In this paper,the basic principle of Radon transform in computer tomography is studied,including the forward and inversion algorithms of tomography.The forward algorithm is based on the Linear Interpolated Ray Tracing Method(LTI),which is a method of dividing a model into rectangular elements and then calculating the ray path according to Fermat's principle.This article implements the process of forward and backward processing.The inversion algorithm uses a joint iterative reconstruction algorithm(SIRT),which obtains the correction value of the grid,which is obtained by averaging the correction values of all the rays passing through the grid,which is easy to cause detail errors.Because if the ray is long,the more plaids that pass,the lower the reliability;if the ray is shorter,the fewer the number of plaids passed,the higher the reliability.Therefore,the SIRT algorithm is improved in this paper,that is,after all the ray-to-grid correction values are found,the weight coefficient is multiplied by the total number of grids through which the ray passes,thereby improving the reconstruction precision.Then,two methods of image reconstruction using Radon transform and image reconstruction using filtered back projection reconstruction algorithm are studied through two experiments.Finally,according to the method of cross-well seismic tomography inversion in geophysics,a method of detecting the internal structure of an object by means of acoustic wave transmission is proposed.The experimental simulation of two different shapes of medium is carried out.At the same time,the influence of the number of iterations on tomographic reconstruction is analyzed.Experiments show that the image reconstruction can be reconstructed by using Radon transform for image reconstruction and filtering back projection reconstruction algorithm for image reconstruction.However,the traditional medical B-mode ultrasound uses sound wave reflection to detect,so that only the surface image of the object is obtained,which is not sensitive to the abnormal reflection in the organ.The LTI-SIRT algorithm uses sound wave transmission to detect the internal details of the object and is less harmful to the human body.At the same time,the method can also be used for the detection of large-scale geological bodies.This method can also be used if the volume of the object to be detected is too large(such as the detection of internal structures of the mountain)or if the depth is too large(such as cross-well seismic detection).The experimental results show that the proposed method can reconstruct the original image well,which indicates that the method can be used for parameter inversion and image reconstruction of complex media.