Multicriteria image space iterative reconstruction

Two-criteria image reconstruction was studied through simultaneous minimization of image peakedness function and sum of squared error between original projection data and reprojection data of reconstructed image. The two-objective optimization can be investigated by the weighted sum scalarization of the two-objective functions. Therefore this is a multicriteria regularization method. The novel multicriteria image space iterative reconstruction overcomes the slow convergence and poor quality of reconstructed image resulted from single-criterion SIRT algorithm, and reduces the pepper-and-salt noise in the image reconstructed by ART algorithm based on minimization of image peakedness function. The image reconstructed by the novel multicriteria image reconstruction algorithm has a good smoothness and fewer noise artifacts. This algorithm can converge more rapidly and steadily to the only optimal result.