Table 3
Sensitivity analysis of triangular membership function parameters.
| Parameter configuration | PSNR (dB) | SSIM | Fuzzy degree adjustment efficiency (%) | Stability score (0–10) | Key observations |
|---|---|---|---|---|---|
| Baseline (Original Parameters) | 32.79 | 0.926 | 91.3 | 8.9 | Default FLS settings with triangular membership functions (slope = 0.5, base width = 0.4, offset = 0.1). |
| High Slope (Slope = 0.8) | 31.42 | 0.908 | 85.6 | 7.3 | Increased slope causes over-sensitivity to input features, leading to unstable fuzzy degree adjustments. |
| Low Slope (Slope = 0.2) | 30.95 | 0.892 | 82.1 | 6.8 | Reduced slope flattens membership curves, resulting in insufficient differentiation between regions. |
| Wide Base (Base Width = 0.6) | 31.84 | 0.912 | 87.4 | 8.1 | Wider base increases overlap between fuzzy sets, reducing specificity. |
| Narrow Base (Base Width = 0.2) | 32.15 | 0.917 | 89 | 8.5 | Narrow base improves regional discrimination but risks overfitting to local features. |
| Offset Shift (Offset = 0.3) | 31.07 | 0.895 | 83.9 | 7 | Excessive offset disrupts alignment between input features and fuzzy sets. |
| No FLS (Ablated) | 31.58 | 0.91 | – | – | Removing FLS entirely degrades performance by 1.21 dB (PSNR) and 0.016 (SSIM). |
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.