Hyperspectral Estimation of Soil Nutrient Content in the Black Soil Region Based on BA-Adaboost

Fig. 5. Measured and predicted values of element content
(a): Adaboost of OM content; (b): BA-Adaboost of OM content; (c): Adaboost of P content;(d): BA-Adaboost of P content; (e): Adaboost of K content; (f): BA-Adaboost of K content

Download full size

View in Article

Table 1. Statistics of nutrient contents in soil samples
View table
View in Article
Table 1. Statistics of nutrient contents in soil samples
养分元素样本数最小值/
(g·kg^-1) 最大值/
(g·kg^-1) 均值/
(g·kg^-1) 标准差/
(g·kg^-1)
有机质 80 2.138 5.517 3.364 0.535
磷 80 0.452 0.988 0.682 0.087
钾 80 1.875 2.240 2.036 0.062

Table 2. Maximum correlation coefficients and sensitive bands

View table

View in Article

Table 2. Maximum correlation coefficients and sensitive bands

元素		光谱变化形式
元素		R	R'	lg1/R	(lg1/R)'
	最大值	-0.678	0.796	0.656	-0.778
有机质	对应波段	600	1 370	600	1 140
	敏感波段数量	95	85	82	101
	最大值	-0.452	0.637	0.458	-0.688
磷元素	对应波段	440	1 390	440	690
	敏感波段数量	11	38	13	82
	最大值	0.428	-0.436	-0.429	0.553
钾元素	对应波段	450	1 650	700~710	980
	敏感波段数量	23	5	26	51

Table 3. Comparison of estimation accuracies

View table

View in Article

Table 3. Comparison of estimation accuracies

元素	光谱变化形式	预测模型	训练集		验证集
元素	光谱变化形式	预测模型	$R_{c}^{2}$	RMSEC	$R_{p}^{2}$	RMSEP
有机质	一阶微分	Adaboost	0.925	0.140	0.756	0.204
有机质	一阶微分	BA-Adaboost	0.961	0.101	0.864	0.152
磷元素	倒数一阶微分	Adaboost	0.909	0.025	0.695	0.039
磷元素	倒数一阶微分	BA-Adaboost	0.944	0.019	0.808	0.031
钾元素	倒数一阶微分	Adaboost	0.877	0.019	0.651	0.029
钾元素	倒数一阶微分	BA-Adaboost	0.917	0.016	0.761	0.024

Tools

Get Citation

Copy Citation Text

Nan LIN, Hai-qi LIU, Jia-jia YANG, Meng-hong WU, Han-lin LIU. Hyperspectral Estimation of Soil Nutrient Content in the Black Soil Region Based on BA-Adaboost[J]. Spectroscopy and Spectral Analysis, 2020, 40(12): 3825

Download Citation

EndNote(RIS)BibTex Plain Text

Set citation alerts for article

Save article for my favorites