Image reconstruction algorithm for EMT based on modified Tikhonov regularization method

摘要

For Electromagnetic Tomography (EMT) Techniques, image reconstruction is one of the crucial steps, which directly affects the quality of reconstruction and real-time performance of the EMT system. Following analyzing the standard Tikhonov regularization method, an improved Tikhonov method is proposed to alleviate the ill-posed characteristics of the EMT inverse problem. The new algorithm involves two steps. First, two weight matrices are estimated using the solution of the standard Tikhonov regularization method, then the image is reconstructed via the improved algorithm proposed in this paper. Experimental tests confirm that the quality of the reconstructed images using the improved algorithm is enhanced to some extent, and iterative procedure is unnecessary, thus the real-time performance can be ensured.