期刊名称：International Journal of Computer Science and Engineering
印刷版ISSN：2278-9960
电子版ISSN：2278-9979
出版年度：2020
卷号：7
期号：2
页码：28-41
DOI：10.14445/23488387/IJCSE-V7I2P104
出版社：IASET Journals
摘要：Image Reconstruction Techniques (IRTs) has been conceded using various reconstruction algorithms. Compared to Analytical image reconstruction method, Statistical image reconstruction methods best suites to reconstruct a high quality image. However, time complexity is involved in it. To overcome the time complexity Maximum Likelihood Expectation Maximization (MLEM) algorithm is parallelized in a multi-core environment. This work concentrates on parallelizing MLEM to reconstruct an image on a shared memory environment in order to reduce the reconstructing time. An attempt is made to optimize the Iteration to reconstruct an image. The performance analyses are employed to know the timeliness, speedup and efficiency for both Sequential and Parallel MLEM. Phantom data set of various sizes under different number of projections is used in our present study. The research shows that the multi-core environment provides the source of high computational power leading to reconstruct an image promptly.
关键词：Image Processing; Image Reconstruction; Iterative Image Reconstruction; Maximum Likelihood Expectation Maximization; Parallel Processing; Open MP
