Abstract: A point spread function and a cost function are obtained with a physical analysis of the high-energy x-ray radiography. Taking the French Test Object model as an example, the conjugate gradient algorithm is applied to the density reconstruction in high-energy x-ray radiography, and the result is satisfactory. The algorithm starts at a simulation in radiography and searches for a maximum likelihood by comparing the simulated radiographs with the measured radiographs. To some extent, the algorithm overcomes the uncertainty in eliminating the blurting effects through a deconvolution process in reconstruction methods.

Key words: x-ray radiography, point spread function, cost function, conjugate gradient algorithm, density reconstruction