Table 4 Comparison of test results of different models27.
Data set type | Method | Recall rate | Accuracy |
|---|---|---|---|
Straight set | ENet | 75.5% | 73.4% |
SCNN | 95.2% | 93.2% | |
Deeplab v3+ | 95.4% | 93.4% | |
Deeplab V3 + network | 95.8% | 94.8% | |
ENet | 71.2% | 70.5% | |
Corner set | SCNN | 94% | 93.5% |
Deeplab v3+ | 94.2% | 93% | |
Deeplab V3 + network | 95.5% | 94.5% | |
Worn lane set | ENet | 70.2% | 69.5% |
SCNN | 93.8% | 91.5% | |
Deeplab v3+ | 92.8% | 90.6% | |
Deeplab V3 + network | 93.5% | 92.8% | |
Night environment | ENet | 40.5% | 38% |
SCNN | 70.5% | 68.1% | |
Deeplab v3+ | 71.2% | 65% | |
Deeplab V3 + network | 65.5% | 62.4% |
