The Research Of Algebraic Reconstruction Algorithm Based On Wavelet Denoising

Computed tomography (CT) is a technique for imaging cross-sections of an object using a series of x-ray measurements in different views, currently, it is the hotspot problem in the field of image processing. Many popular image reconstruction algorithms are researched and summarized in this dissertation. In this paper,image denoising technique based on wavelet transform and tomography algorithm with incomplete projection data are mainly studied.Firstly, a new threshold denoising algorithm based on wavelet transform has been put forward, by researching and analysis of the existing denoising arithmetics which have larger interference from external environment in the process of collect the projection data. For achieving effect of removing noise while preserving image details efficiently, a new threshold function is put forward in this algorithm by the features distribution for image signal and noise, and introduce an improved method of threshold selection by a singular factor combine with the abundance extent of the image texture and edge.Secondly, a improved algebraic reconstruction algorithm is put forward according to the problem of large calculation and large thereserves. For reduceing the capacity of the coefficient matrix, the coefficient matrix is decomposited for the special matrix with ray and matrix with coefficient for single ray to avoid duplication of storage. In the reconstruction process, the power factor is simplified in order to improve the reconstruction speed, error is exist, so a method to improve the reconstruction quality is proposed, smoothing the reconstructed image to obtain optimum effect of the reconstruction.Finally, I download the datum of image processing from image processing analysis web site and simulate arithmetics above in Matlab. The experiment results show that the proposed algorithm is effective and practical.