Research On Algorithm For CT Iterative Reconstruction Based On Variable Voltage

For the projection image sequence obtained by the variable voltage CT imaging system,it’s easy to appear the phenomenon of underexposed and overexposed. The projection information lacks badly which impacts the quality of the CT reconstruction. The method based on gray weighted can get complete structural information of the complex workpieces,but it’s easy to make the reconstructed results lose the details through linear optimization in the process of weighted fusion. To this issue, for the characteristics that the image sequence on variable voltage is locally effective, the iterative reconstruction algorithm is used to reconstruct recursively effective area under variable voltage. Taking into account the impact of low doses of noise on low voltages, the sparsity of image and the structural information of image sequence on variable voltage are used as a priori knowledge to apply to the CT iterative reconstruction algorithm. Based on this, the CT iterative reconstruction algorithms on variable voltage based on a priori information are studied in this dissertation, which improves the quality of the reconstructed results.Firstly, an algorithm for variable voltage CT iterative reconstruction based on prior information is proposed on the basis of the principle of variable voltage CT imaging. In the process of reconstruction, the reconstructed result which is obtained by using TV-ART to reconstruct the effective projection region on the low voltage is used as an initial value of the reconstruction of the adjacent high voltage, to the highest voltage according to the algorithm.That is to say the structural information is completely reconstructed. And the feasibility of the method is verified by simulation experiment. Then, a method for image restoration based on prior structure information is put forward for the problem of edge degradation by using variable voltage TV-ART iterative algorithm, which improves the quality of the reconstructed result of variable voltage TV-ART iterative algorithm. Finally, for the poor quality problem of projection data because of noise, a variable voltage TV-EM iterative algorithm based on priorinformation is advanced, which suppresses noise and artifacts in the reconstruction result and to further improves the quality of the reconstruction images on variable voltage. The result of the simulation and the practical experiments validate the effectiveness of the method.