Local Search Method for Image Reconstruction with Same Concentration in Tomography

Image reconstruction in tomography is an attractive research area that has received considerable attention in recent years. The image reconstruction can be viewed as an optimization problem where the main objective is to obtain high quality reconstructed images. In this paper, we proposed a local search (LS) method to improve the quality of reconstructed images in tomography in supposed case of similar concentration of physical phenomena. The proposed method starts with an initial image solution and tries to enhance its quality. A solution is a set of points where each point represents a distribution of a physical quantity resulting by radius emission. Each point is evaluated by a function that estimates the difference between the estimated and the measured projections. The LS makes use of a move operator that permits to generate neighbour solutions and helps in finding the optimal correctness of distribution in each point. The LS is an iterative process that tries to optimize position of the physical parameter on the image in order to obtain a solution corresponding to the reconstructed image. To measure the performance of the proposed approach, we have implemented it and compared it with the Filtered back-projection (FBP). Further, we compared the reconstructed images of LS with the source ones. The numerical results are promising and demonstrate the benefits of the proposed approach.