Based On The Sparse Point Of CT Image Reconstruction Algorithm Of ROF Model

Computed tomography(CT) has the advantages of high density and non destructive testing, etc. So it is widely used in the fields of medical diagnosis, industrial non-destructive testing and safety inspection. Because of the objective conditions such as radiation dose, imaging hardware limitations and so on, the projection data are often incomplete projection data in actual application. Sparse angle projection is often used to obtain incomplete projection data in medical imaging. It is difficult to accurately reconstruct the CT image using the analytical method represented by filtered back projection(FBP). Therefore, the study of sparse angle incomplete projection data imaging, it has far-reaching significance for the development of CT system.Firstly, this paper expounded the theoretical basis of CT image reconstruction, introduced the principle of CT image reconstruction and mathematical basis. Secondly, the two common methods of CT image reconstruction which are analytical method and iterative method were introduced. Three groups of experiments were conducted, they analyzed the characteristics of FBP, SART, SIRT and SART algorithms reconstruction and demonstrate the SART algorithm which was used to reconstruct the sparse angle CT image, especially the CT image with noise is the best. Thirdly, because the SART algorithm is limited in the process of CT image reconstruction, we proposed a new method based on ROF and Laplace model and demonstrated it can not only remove the noise effectively, but also keep the edge of the image better. Finally, the ROF-LP model often took a long time and the result image was easy to produce the "step effect". So the wavelet transform was introduced. Let reconstruction and de-noising process use alternating manner. Threshold denoising and Denoising using ROF-LP model for high frequency part of wavelet transform, enhancing coefficients for low frequency part at the same time. The experimental results show that this method can speed image reconstruction, suppress the generation of the staircase effect and improve the quality of the reconstructed image...