ALTERNATIVE ERROR MEASURES FOR NON-DIFFRACTION

Most algebraic reconstruction techniques for non-diffraction and diffraction tomography are based on minimization of the 12 no~ residual error. We investigate the properties and characteristics of tomographic reconstructions based on the 11 no~ using both Simplex and Karmarkar algorithms as well as the Ip no~s for 1least squares, and results are compared to 12-norm reconstructions. We also consider a total least squares error criterion. I. INTRODUCTION Tomographic image reconstruction has found important applications in several areas of inquiry, including medical diagnosis and nondestructive testing. Recently, attention has been focused on the possible applications of tomographic image reconstruction in geophysics.