Parallel Image Reconstruction Using the Maximum Likelihood Method with a Graphics Processor and the OpenGL Library

Abstract

Creating fast parallel iterative statistical algorithms based on the use of graphics accelerators is an important and urgent task of great scientific and practical importance. An algorithm based on the method of maximizing the maximum likelihood expectation (maximum likelihood expectation maximization—MLEM) is considered. The MLEM is a numerical method for determining maximum likelihood estimates and, since its first application in the field of image reconstruction in 1982, remains one of the most popular statistical image reconstruction methods and is the foundation for many other approaches. A new version of the MLEM parallel algorithm is proposed that provides global convergence of the iterative algorithm. To parallelize the algorithm, we use the texture mapping method using the OpenGL graphics library. The parallel algorithm is described in as much detail as possible. Examples of several reconstructions of images of aluminum casting products are given. The obtained result can be used for nondestructive testing of various industrial products, including testing of foundry products.