Table 2 Quantitative analysis results of five models on the deepglobe road dataset.

From: A dynamic attention mechanism for road extraction from high-resolution remote sensing imagery using feature fusion

Scheme

Network

Precision (%)

Rcall (%)

IoU (%)

F1-Score (%)

(a)

UNet

88.81

75.26

77.80

81.64

(b)

KanUNet

89.75

77.73

76.03

83.26

(c)

DenseUNet

91.32

80.02

81.75

85.32

(d)

MDCGANet

91.68

82.96

77.72

87.12

(e)

SCSM

92.03

81.35

80.41

86.51

(f)

RISENet

92.24

85.72

82.77

88.86

  1. Significant values are in [bold].
Back to article page