Table 13 Environmental corruption impact on UAV tracking performance.
From: Advanced algorithms for UAV tracking of targets exhibiting start-stop and irregular motion
Corruption Type | Standard KF | OC-SORT | SFTrack | UAVMOT + AMF | Flow-Guided | SMART-TRACK |
|---|---|---|---|---|---|---|
Weather Corruptions | ||||||
Fog | 32.1% | 41.3% | 48.7% | 45.2% | 52.4% | 49.1% |
Rain | 45.8% | 52.6% | 58.3% | 56.7% | 61.2% | 59.4% |
Snow | 38.4% | 46.9% | 53.1% | 50.8% | 56.7% | 54.3% |
Frost | 41.2% | 48.7% | 54.9% | 52.3% | 58.1% | 55.6% |
Sensor Corruptions | ||||||
Gaussian Noise | 35.7% | 44.2% | 50.8% | 47.6% | 53.9% | 51.2% |
Impulse Noise | 33.9% | 42.1% | 48.6% | 45.8% | 52.1% | 49.7% |
Low Contrast | 31.4% | 40.6% | 47.2% | 44.3% | 50.8% | 48.1% |
Blur Corruptions | ||||||
Motion Blur | 29.8% | 38.9% | 45.3% | 42.7% | 49.1% | 46.8% |
Defocus Blur | 36.2% | 44.8% | 51.4% | 48.9% | 54.6% | 52.1% |
Composite Corruptions | ||||||
Rain-Defocus | 28.5% | 37.2% | 43.9% | 41.1% | 47.6% | 45.2% |
Rain-Gaussian | 22.3% | 31.8% | 38.4% | 35.7% | 42.1% | 39.6% |
Average Performance | 34.1% | 42.6% | 49.1% | 46.4% | 52.3% | 50.1% |
