Table 7 Comparison of predicted results with different methods.

From: LapUNet: a novel approach to monocular depth estimation using dynamic laplacian residual U-shape networks

Ranging point

Truth value/m

Estimated value/m

Absolute error/m

Relative error/%

Lapdepth

Monodepth2

LapUNet

Lapdepth

Monodepth2

LapUNet

Lapdepth

Monodepth2

LapUNet

A1

9.11

9.62

9.45

9.31

0.51

0.34

0.2

5.6%

3.7%

2.2%

A2

14.75

16.15

13.13

15.18

1.40

1.62

0.43

9.5%

10.9%

2.9%

A3

19.62

18.56

21.27

20.63

1.06

1.65

1.01

5.4%

8.4%

5.1%

B1

8.92

8.70

8.64

8.65

0.22

0.28

0.27

2.5%

3.1%

3.0%

B2

13.58

16.07

15.58

13.96

2.51

2.00

0.38

18.3%

14.7%

2.7%

B3

18.87

21.27

21.83

19.98

2.40

2.96

1.11

12.7%

15.7%

5.8%

B4

24.57

24.91

25.57

25.97

0.34

1.00

1.4

1.4%

4.1%

5.6%

C1

9.61

9.31

9.69

9.31

0.30

0.08

0.3

3.1%

0.8%

3.1%

C2

13.99

15.78

13.92

14.25

1.79

0.07

0.26

12.8%

0.5%

1.8%

C3

19.56

21.41

20.48

19.96

1.85

0.92

0.4

9.5%

4.7%

2.0%

D1

8.53

9.22

8.74

8.47

0.69

0.21

0.06

8.1%

2.5%

0.7%

D2

14.81

14.16

14.15

14.32

0.65

0.66

0.49

4.4%

4.5%

3.3%

Back to article page