An algorithm is proposed that can reconstruct the image with better quality from projection data containing this kind of noise. The effectiveness of the proposed method is verified by computer simulation. In this method, the portion with high SN ratio (i. e. , high reliability) in the obtained projection data is emphasized and effectively utilized, while the contribution of the portion of projection data with low SN ratio and low reliability is suppressed in order to reduce the effect of quantum noise. The reconstruction is made by an iterative computation technique. As a result of simulation, it is seen that, by suitable choice of the weights of the projection data base on the reliability of the measure data and by suitable setting of the terminating condition for the iteration, the effect of the quantum noise is decreased, leading to a reconstructed image of satisfactory quality.
|Number of pages||9|
|Journal||Systems, computers, controls|
|Publication status||Published - 1984 Jan 1|
ASJC Scopus subject areas