THE MOVING VOXEL IMAGE RECONSTRUCTION (MVIR) METHOD FOR LANE DETERMINATION IN MULTILANE PORTAL MONITOR SYSTEMS

摘要

Radiation portal monitor stations can potentially be used at locations where vehicles travel at relatively high speeds. These are more challenging environments than those where vehicles stop or drive through slowly. We have developed a dynamic image reconstruction method for use in such multilane EZ-Pass portals. Called MVIR (Moving Voxel Image Reconstruction), this method detects the lane and time at which radioactive sources pass through with significantly better accuracy and precision than is possible using conventional static image reconstruction methods. Rather than producing an image over a static (x, y) grid, MVIR combines consecutive time slices of data into a single linear system of equations that is solved for the source intensity as a function of time and position. This improves counting statistics and enables a small number of detectors to produce relatively high resolution images. In this presentation we compare the results of image reconstructions using MVIR with a static image reconstruction method, on data simulated using the Los Alamos PortalSim software.