Hyperspectral Estimation of Wheat Leaf Water Content Using Fractional Differentials and Successive Projection Algorithm-Back Propagation Neural Network

Hasan Umut; Sawut Mamat; Chunyue Ma

doi:10.3788/LOP56.153002

Journals >Laser & Optoelectronics Progress >Volume 56 >Issue 15 >Page 153002 > Article

Laser & Optoelectronics Progress
Vol. 56, Issue 15, 153002 (2019)

Hyperspectral Estimation of Wheat Leaf Water Content Using Fractional Differentials and Successive Projection Algorithm-Back Propagation Neural Network

Hasan Umut^1、2, Sawut Mamat^{1、2、3、*}, and Chunyue Ma^1、2

Author Affiliations

¹ College of Resource and Environment Sciences, Xinjiang University, Urumqi, Xinjiang 830046, China

² Key Laboratory of Oasis Ecology of Ministry of Education, Urumqi, Xinjiang 830046, China

³ Key Laboratory for Wisdom City and Environmental Modeling, Xinjiang University, Urumqi, Xinjiang 830046, China

show less

DOI: 10.3788/LOP56.153002 Cite this Article Set citation alerts

Hasan Umut, Sawut Mamat, Chunyue Ma. Hyperspectral Estimation of Wheat Leaf Water Content Using Fractional Differentials and Successive Projection Algorithm-Back Propagation Neural Network[J]. Laser & Optoelectronics Progress, 2019, 56(15): 153002 Copy Citation Text

show less

Canopy spectral curves of spring wheat. (a) Canopy spectral curves of spring wheat with different water contents; (b) canopy spectral curves of spring wheat with 0-order to 2-order differentials

Fig. 1. Canopy spectral curves of spring wheat. (a) Canopy spectral curves of spring wheat with different water contents; (b) canopy spectral curves of spring wheat with 0-order to 2-order differentials

Download full size

Correlation coef?cients between MLWC and spectral re?ectance. (a) Correlation coefficients between MLWC and 0-order spectrum; (b) correlation coefficients between MLWC and 0.2, 0.4, 0.6, 0.8-order spectra; (c) correlation coefficients between MLWC and 1-order spectrum; (d) correlation coefficients between MLWC and 1.2-order spectrum; (e) correlation coefficients between MLWC and 1.4-order spectrum; (f) correlation coefficients between MLWC and 1.6-order spectrum; (g) correlation coefficients bet

Fig. 2. Correlation coef?cients between MLWC and spectral re?ectance. (a) Correlation coefficients between MLWC and 0-order spectrum; (b) correlation coefficients between MLWC and 0.2, 0.4, 0.6, 0.8-order spectra; (c) correlation coefficients between MLWC and 1-order spectrum; (d) correlation coefficients between MLWC and 1.2-order spectrum; (e) correlation coefficients between MLWC and 1.4-order spectrum; (f) correlation coefficients between MLWC and 1.6-order spectrum; (g) correlation coefficients bet

Download full size

Fig. 3. Fitting analysis results between measured values and predicted values by BP neural network model. (a) 1-order differential; (b) 1.2-order differential; (c) 1.4-order differential; (d) 1.6-order differential; (e) 1.8-order differential; (f) 2-order differentials

Download full size

Differentialorder	Numberof bands	Band combinationsselected by SPA/nm
0	0	—
0.2	0	—
0.4	0	—
0.6	0	—
0.8	0	—
1	2	867、2089
1.2	13	420,486,616,690,912,957,1223,1660,1978,2103,2111,2238,2276
1.4	4	859,888,1255,1953
1.6	5	420,1278,1627,2238,2286
1.8	6	823,875,892,1004,1627,2286
2	5	888,911,1144,1686,2286

Table 1. Numbers and combinations of bands selected by SPA

Differentialorder	Optimal BPneuralnetworkstructure	θ_C	$\begin{matrix} R_{C}^{2} \end{matrix}$	θ_V	$\begin{matrix} R_{V}^{2} \end{matrix}$	δ
1	2-4-1	0.692	0.737	0.496	0.628	1.395
1.2	13-12-1	1.008	0.472	0.366	0.789	2.050
1.4	4-5-1	1.174	0.482	0.417	0.809	1.943
1.6	5-9-1	0.553	0.828	0.634	0.807	1.495
1.8	6-4-1	0.701	0.751	0.227	0.917	3.253
2	5-8-1	0.733	0.698	0.308	0.840	2.298

Table 2. Comparison of modeling results

Download Citation

Set citation alerts for the article

Tools

Set citation alerts for the article

Save the article for my favorites

Paper Information