The Comparison of Three Reconstruction Methods: Straight Line, Point of the Closest Approach (POCA), Maximum Likelihood Expectation Maximization (MLEM)

Abstract

Muon tomography offers an effective opportunity to image the inner parts of big and heavy objects by using cosmic ray muons. This is a comparative study that presents the performance of the three reconstruction instruments: Point of Closest Approach (POCA), Straight Line, and Maximum Likelihood Expectation-Maximization (MLEM) in muon tomography. The efficacy and accuracy of these instruments are assessed by simulating data that is the resulting effect of the GEANT4 toolkit. The setup of the simulation is close to the real experimental conditions: six virtual sensors are set above and beneath the object to pick the trajectory of the muons. The POCA technique gives the assignment of the scattering centers by computing the points of closest approach for the incoming and outgoing muons. The Straight Line technique assumes that the muon paths are 2 linear. It selects the point in the middle of the last incoming and the first outgoing points. The MLEM technique is iterative and refines an image iteratively by maximizing the likelihood of the observed data. While the POCA method and the Straight Line are methods for efficient computation, the MLEM is more accurate than they are, in particular when regarding the application in media that are this heterogeneous. It will display the pros and cons of each method and help in identifying the ideal muon tomography processes, which will cover archaeological investigations, nuclear security, and industrial inspection.