Skip to main content

Table 7 Summary of the LMMs, created according to the formulas LMM[1a]-LMM[3d], and the reference linear models (LMs), to predict TLS-derived horizontal crown projection (CPATLS) with SIP-derived data, showing model prediction errors: RMSE1 (based on model fitted values), RMSE2 (based on a leave-one-out analysis); the best predictive model is denoted by italics; and all models’ coefficients are given in the Electronic supplementary material

From: Single-image photogrammetry for deriving tree architectural traits in mature forest stands: a comparison with terrestrial laser scanning

Model ID

RMSE1 (m2)

RMSE1 (%)

RMSE2 (m2)

RMSE2 (%)

LM[1]

9.94

27.59

10.23

28.37

LMM[1a]

8.27

22.94

9.49

26.32

LMM[1b]

8.25

22.89

9.90

27.48

LM[2]

8.99

24.96

9.56

26.51

LMM[2a]

7.75

21.50

9.16

25.43

LMM[2b]

7.48

20.76

8.97

24.90

LMM[2c]

5.67

15.75

7.45

20.66

LM[3]

8.65

24.00

9.59

26.61

LMM[3a]

7.79

21.62

9.44

26.20

LMM[3b]

7.53

20.90

9.28

25.75

LMM[3c]

5.68

15.76

7.82

21.69

LMM[3d]

7.63

21.18

9.17

25.43