Table 1 Average X-, Y-, and Z-spatial errors for T-PEPT, PEPT-ML, and the Birmingham Method for the 600 RPM case scenarios. Data for which PEPT-ML and the Birmingham Method were unable to locate all 5 tracers successfully are not considered. The number of LORs per sample was 1000 for all three algorithms.
Scenario | Method | No. tracers located | Error (mm) | ||
|---|---|---|---|---|---|
X | Y | Z | |||
Glass tracer, aluminium walls | T-PEPT | 5 | 0.64 (15) | 0.97 (15) | 0.67 (15) |
PEPT-ML | 5 | 0.83 (14) | 1.16 (14) | 0.91 (14) | |
BHAM Method | 4 | 1.2 (3) | 2.0 (3) | 1.4 (3) | |
Glass tracer, steel walls | T-PEPT | 5 | 1.84 (19) | 2.28 (19) | 1.94 (19) |
PEPT-ML | 5 | 2.25 (19) | 2.68 (19) | 2.31 (19) | |
BHAM Method | 4 | 3.8 (4) | 4.6 (4) | 3.8 (4) | |
Steel tracer, steel walls | T-PEPT | 5 | 2.56 (14) | 2.90 (14) | 2.66 (14) |
PEPT-ML | 4 | 4.2 (4) | 5.1 (4) | 4.3 (4) | |
BHAM Method | 3 | 7.4 (8) | 9.2 (8) | 7.9 (8) | |
