Inversion of Wheat Tiller Density Based on Visible-Band Images of Drone

Meng-meng DU; Roshanianfard Ali; Ying-chao LIU; [in Chinese]

doi:10.3964/j.issn.1000-0593(2021)12-3828-09

Journals >Spectroscopy and Spectral Analysis >Volume 41 >Issue 12 >Page 3828 > Article

Spectroscopy and Spectral Analysis
Vol. 41, Issue 12, 3828 (2021)

Inversion of Wheat Tiller Density Based on Visible-Band Images of Drone

Meng-meng DU^1、*, Roshanianfard Ali^2、2;, Ying-chao LIU^3、3;, and [in Chinese]^2、2;

Author Affiliations

¹1. School of Agricultural Equipment Engineering, Henan University of Science and Technology, Luoyang 471003, China

²2. Faculty of Agriculture and Natural Resources, University of Mohaghegh Ardabili, 56199-11367, Ardabil, Iran

³3. School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

show less

DOI: 10.3964/j.issn.1000-0593(2021)12-3828-09 Cite this Article

Meng-meng DU, Roshanianfard Ali, Ying-chao LIU, [in Chinese]. Inversion of Wheat Tiller Density Based on Visible-Band Images of Drone[J]. Spectroscopy and Spectral Analysis, 2021, 41(12): 3828 Copy Citation Text

show less

Study area and positions of ground truths of wheat tiller density(a): Experimental field; (b): Experimental field and positions of ground truths of wheat tiller density;(c) Counting wheat tiller numbers within the zone of “1 meter and 2 rows”

Fig. 1. Study area and positions of ground truths of wheat tiller density
(a): Experimental field; (b): Experimental field and positions of ground truths of wheat tiller density;(c) Counting wheat tiller numbers within the zone of “1 meter and 2 rows”

Download full size | View in the Article

Fig. 2. Calibrated drone remote sensing images of blue, green and red bands
(a): Blue band; (b): Green band; (c): Red band

Download full size | View in the Article

Fig. 3. Spectral response characteristics of vegetation and soil in visible light domain

Download full size | View in the Article

Fig. 4. Training model of BP neural network

Download full size | View in the Article

Fig. 5. Vegetation index maps

Download full size | View in the Article

Fig. 6. Training result of BP neural network model

Download full size | View in the Article

Fig. 7. Map of quantitatively inversed wheat tiller density at field level

Download full size | View in the Article

	项目	值
大疆Mini 2 无人机	外形尺寸/mm	245×289×56
	净重/g	249
	最大爬升速度/(m·s^-1)	5
	最大水平飞行速度/(m·s^-1)	16
相机模块	视角/(°)	83
	传感器类别	CMOS
	有效像素	3 000×4 000
	等效焦距/mm	24
	图片格式	JPEG/DNG (RAW)

Table 1. Experiment equipment

View in the Article

序号	“1米双行” 小麦茎蘖数	小麦茎蘖密度地面真值/(s·m^-2)	FVC 平均值	VDVI 平均值	NGRDI 平均值	NGBDI 平均值	RGRI 平均值
1	138	690	0.483	0.152	0.075	0.242	1.166
2	173	865	0.571	0.193	0.115	0.283	1.269
3	137	685	0.472	0.147	0.068	0.238	1.150
4	153	765	0.501	0.16	0.082	0.252	1.184
5	136	680	0.475	0.149	0.072	0.237	1.168
6	126	630	0.467	0.145	0.065	0.239	1.140
7	145	725	0.498	0.159	0.082	0.248	1.185
8	116	580	0.430	0.128	0.048	0.222	1.103
9	91	455	0.402	0.115	0.039	0.203	1.086
10	162	810	0.518	0.168	0.086	0.264	1.191
11	102	510	0.422	0.124	0.045	0.217	1.098
12	163	815	0.540	0.178	0.098	0.272	1.220
13	105	525	0.422	0.124	0.046	0.216	1.099
14	110	550	0.433	0.129	0.05	0.222	1.111
15	148	740	0.518	0.168	0.088	0.263	1.199
16	94	470	0.493	0.157	0.078	0.249	1.173
17	89	445	0.418	0.123	0.044	0.214	1.094
18	85	425	0.456	0.140	0.058	0.237	1.129
19	170	850	0.542	0.179	0.102	0.268	1.234
20	97	485	0.481	0.151	0.072	0.243	1.161

Table 2. Ground truths of wheat tiller density and corresponding VI values

View in the Article

VI	植被				土壤
VI	最大值	最小值	均值	标准差	最大值	最小值	均值	标准差
FVC	0.851	0.687	0.770	0.052	0.184	0.133	0.158	0.016
VDVI	0.544	0.301	0.389	0.071	-0.013	-0.256	-0.069	0.089
NGRDI	0.250	0.167	0.204	0.026	-0.035	-0.073	-0.047	0.011
NGBDI	0.474	0.323	0.375	0.043	0.072	0.046	0.059	0.008
RGRI	1.667	1.400	1.516	0.083	0.932	0.863	0.911	0.020

Table 3. Statistic values of vegetation and soil in FVC, VDVI, NGRDI, NGBDI and RGRI maps

View in the Article

序号	小麦茎蘖密度地面真值 /(株·m^-2)	小麦茎蘖密度预测值 /(株·m^-2)
1	470	455
2	445	455
3	425	455
4	850	865
5	485	465

Table 4. Ground truths and prediced values of wheat tiller density

Meng-meng DU, Roshanianfard Ali, Ying-chao LIU, [in Chinese]. Inversion of Wheat Tiller Density Based on Visible-Band Images of Drone[J]. Spectroscopy and Spectral Analysis, 2021, 41(12): 3828

Download Citation

Tools

Save the article for my favorites

Paper Information