Advances, Systems and Applications
Table 5 Influence of metrics in the balanced VQA model
From: Automated visual quality assessment for virtual and augmented reality based digital twins
Metric | Influence (%) |
|---|---|
Surface area ratio | \(7.48 \pm 0.45\) |
Normal deviation, 1.0, mean | \(4.45 \pm 0.40\) |
Matching distance, 1.0, mean | \(4.33 \pm 0.41\) |
Normal deviation, 1.5, mean | \(3.05 \pm 0.40\) |
Matching distance, 1.5, mean | \(2.88 \pm 0.35\) |
Matching distance, 2.0, mean | \(2.41 \pm 0.25\) |
Polygon area COV ratio | \(2.04 \pm 0.21\) |
Convexity ratio | \(2.03 \pm 0.10\) |
Matching distance, 2.5, mean | \(2.00 \pm 0.20\) |
Test mesh bounding box density | \(1.97 \pm 0.10\) |
Sphericity ratio | \(1.92 \pm 0.19\) |
Shape efficiency ratio | \(1.86 \pm 0.09\) |
Normal deviation, 1.0, minimum | \(1.85 \pm 0.17\) |
Normal deviation, 2.0, mean | \(1.73 \pm 0.25\) |
Matching distance, 3.0, mean | \(1.49 \pm 0.15\) |
Dihedral angle change, 1.0, COV | \(1.42 \pm 0.21\) |
Normal deviation, 2.5, mean | \(1.41 \pm 0.06\) |
Polygon area skewness ratio | \(1.36 \pm 0.16\) |