Depth Inversion of Coastal Waters Based on Bi-LSTM

Band number	Band	Central wavelength /μm	Spatial resolution /m
1	Coastal aerosol	0.443	60
2	Blue	0.490	10
3	Green	0.560	10
4	Red	0.665	10
5	Vegetation red edge	0.705	20
6	Vegetation red edge	0.740	20
7	Vegetation red edge	0.783	20
8	NIR	0.842	10
8a	Vegetation red edge	0.865	20
9	Water vapour	0.945	60
10	SWIR-Cirrus	1.375	60
11	SWIR	1.610	20
12	SWIR	2.190	20

Table 2. Accuracy comparison of different models
View table
Table 2. Accuracy comparison of different models
Model MRE /% RMSE /m MAE /m
Bi-LSTM 18.93 0.85 0.60
1D-CNN 33.12 1.04 0.76
Improved Stumpf 37.44 0.97 0.76
Stumpf 47.99 1.12 0.88

Table 3. Accuracy comparison of different models
View table
Table 3. Accuracy comparison of different models
Model MRE /% RMSE /m MAE /m R²
Bi-LSTM 18.44 0.88 0.63 0.93
1D-CNN 32.41 1.10 0.82 0.90
ImprovedStumpf 37.49 0.98 0.77 0.91
Stumpf 45.08 1.12 0.89 0.89

Table 4. Accuracy comparison of different models under different water depth intervals

View table

Table 4. Accuracy comparison of different models under different water depth intervals

Model	Evaluation method	0--2 m	2--4 m	4--6 m	6--8 m	8--10 m	10--12 m
	MRE /%	208.17	29.63	22.49	11.38	7.61	8.62
Stumpf	RMSE /m	1.30	1.02	1.38	0.97	0.86	1.09
	MAE /m	0.97	0.84	1.12	0.78	0.68	0.90
	MRE /%	157.65	22.30	19.16	9.78	8.39	7.33
Improved Stumpf	RMSE /m	1.07	0.82	1.19	0.84	0.92	0.94
	MAE /m	0.77	0.63	0.96	0.67	0.75	0.76
	MRE /%	124.89	27.09	18.58	13.77	8.70	5.68
1D-CNN	RMSE /m	0.94	1.06	1.15	1.24	0.96	0.87
	MAE /m	0.59	0.74	0.93	0.95	0.77	0.60
	MRE /%	57.92	16.85	14.69	10.48	8.53	5.11
Bi-LSTM	RMSE /m	0.62	0.67	0.99	1.01	0.96	0.76
	MAE /m	0.38	0.48	0.73	0.72	0.76	0.53

Tools

Get Citation

Copy Citation Text

Xinliang Pan, Renhui Yang, Tao Jiang, Baikai Sui, Chenxi Liu, Zhen Zhang. Depth Inversion of Coastal Waters Based on Bi-LSTM[J]. Acta Optica Sinica, 2021, 41(10): 1012003

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites