Spectroscopy and Spectral Analysis, Volume. 41, Issue 12, 3949(2021)
Research on Large-Scale Monitoring of Spider Mite Infestation in Xinjiang Cotton Field Based on Multi-Source Data
Figures & Tables(13)
Fig. 1. Images of cotton leaves with different degree of mite damage
(a): Level 0; (b): Level 1; (c): Level 2; (d): Level 3
Fig. 3. Spectral reflectance curves of ground canopies with different degrees of mite damage
Fig. 4. First derivative spectral curve of ground canopy with the different degrees of spider mites damage
Fig. 5. Reflectance curve of multispectral image with different degrees of spidermite damage
Fig. 6. Spatial distribution map (a, b, c, d, e, f) of cotton field spider mite monitoring in (June 22, 27, 28, Jule 3, 9, 13) different periods
Table 1. Calculation formulas for vegetation index
View tableView in ArticleTable 1. Calculation formulas for vegetation index
植被指数 公式 NDGI (G-R)/(G+R) TVI 0.5×[120×(NIR-G)-200×(R-G)][ 10 ]MSR (NIR/R-1)/[(NIR/R)0.5+1][ 10 ]MSAVI 0.5×{(2×NIR+1)-[(2×NIR+1)2-
8×(NIR-R)]0.5}[11 ]RDVI (NIR-R)/[(NIR+R)0.5][ 10 ]RVI NIR/R[ 11 ]REWDRVI (0.15×NIR-RE)/(0.15×NIR+RE)[ 11 ]DVI NIR-R[ 10 ]GNDVI (NIR-G)/(NIR+G)[ 11 ]SAVI 1.5×(NIR-R)/(NIR+R+0.5)[ 12 ]ARI (1/G)-(1/RE) OSAVI 1.16×(NIR-R)/(NIR+R+0.16)[ 12 ]NDVI (NIR-R)/(NIR+R)[ 12 ]EVI 2.5×[(NIR-R)/(NIR+6×R-7.5×G+1)][ 13 ]GRVI NIR/G[ 13 ]GDVI NIR-G[ 13 ]GOSAVI (1+0.16)×(NIR-G)/(NIR+G+0.16)[ 13 ]REDVI NIR-RE[ 13 ]RERVI NIR/RE[ 13 ]RESAVI 1.5×[(NIR-RE)/(NIR+RE+0.5)][ 13 ]RERDVI (NIR-RE)/[(NIR+RE)0.5] [ 13 ]RENDVI (NIR-RE)/(NIR+RE)[ 13 ]REOSAVI (1+0.16)×(NIR-RE)/(NIR+RE+0.16)[ 13 ]
Table 2. Correlation between the occurrence of spider mite and vegetation indices
View tableView in ArticleTable 2. Correlation between the occurrence of spider mite and vegetation indices
植被指数 RDVI SAVI OSAVI TVI NDGI RVI MSR 相关系数 0.407* 0.222** 0.304** 0.492** -0.304** 0.210* 0.197*
Table 3. Correlation between the occurrence of spider mite and environmental data
View tableView in ArticleTable 3. Correlation between the occurrence of spider mite and environmental data
环境数据 最高温度 平均温度 平均湿度 温湿系数 积温 10 cm土壤最高温度 10 cm土壤平均温度 10 cm土壤平均湿度 相关系数 -0.218* -0.178* 0.221** 0.243** -0.213* -0.201* -0.258** 0.213*
Table 4. Comparison of classification results of the different degrees of spider mite monitoring models
View tableView in ArticleTable 4. Comparison of classification results of the different degrees of spider mite monitoring models
建模方式 样本 训练样本 测试样本 健康 螨害 准确率/% 健康 螨害 准确率/% 精确率/% 召回率/% F1值/% 单一环境数据M1 健康
螨害26
923
3264.44 13
108
1460 56.52 61.9 59.09 单一植被指数M2 健康
螨害27
129
4276.67 15
86
1668.89 65.22 71.43 68.18 环境数据与植被
指数结合M3健康
螨害26
313
4882.22 12
45
2480 75 70.59 72.73
Table 5. Correlation coefficient between environmental data and cotton field spider mite area
View tableView in ArticleTable 5. Correlation coefficient between environmental data and cotton field spider mite area
环境数据 最高温度 平均湿度 温湿系数 积温 10 cm土壤最高温度 10 cm土壤平均温度 相关系数 0.707* 0.844* 0.931** 0.837* 0.856* 0.974**
Table 6. Model evaluation results
Table 7. Comparison of prediction results of cotton field spider mite area prediction models
View tableView in ArticleTable 7. Comparison of prediction results of cotton field spider mite area prediction models
日期 实际值/亩 预测值/亩 6.22 64.168 58.374 6.27 18.361 18.591 6.29 20.348 18.101 7.3 50.055 41.666 7.9 43.087 42.39 7.13 72.831 69.83
Tools
Get Citation
Copy Citation Text
Li-li YANG, Zhen-peng WANG, Cai-cong WU. Research on Large-Scale Monitoring of Spider Mite Infestation in Xinjiang Cotton Field Based on Multi-Source Data[J]. Spectroscopy and Spectral Analysis, 2021, 41(12): 3949
Paper Information
Category: Research Articles
Received: Nov. 3, 2020
Accepted: --
Published Online: Dec. 17, 2021
The Author Email: Li-li YANG (llyang@cau.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20