Monitoring rice bacterial blight based on UAV images of different ground sampling distances (GSD)
-
摘要:目的
快速无损地监测水稻白叶枯并量化感染程度，从而指导田间作业。
方法利用无人机获取受白叶枯病菌Xanthomonas oryzae pv. oryzae感染的水稻冠层高分辨率图像，提取颜色和纹理特征，分别构建基于颜色、纹理特征以及两者融合的多元回归模型，对白叶枯感染等级进行预测。探究不同地面分辨率(Ground sampling distances，GSD)对不同模型精度的影响。
结果基于颜色特征的监测模型的决定系数(Coefficient of determination，R2)为85.9%，均方根误差(Root mean square error，RMSE)为1.43，相对均方根误差(Relative RMSE，RRMSE)为19.1%，相比基于纹理特征的模型，R2上升了2.4个百分点，RRMSE增加了4.6个百分点；与单一种类特征相比，基于颜色和纹理特征融合的预测模型(R2=89.6%，RMSE=1.06，RRMSE=15.1%)精度有较大的提升；通过构建不同GSD模型发现，当GSD为0.2 、0.5或0.8 cm时，模型精度较高，R2 均在80%以上。
结论从无人机捕获的低空遥感图像中提取的颜色和纹理特征可用于监测水稻白叶枯病，结果可为无人机遥感监测水稻白叶枯提供有效的技术支持。
Abstract:ObjectiveIn order to monitor rice bacterial blight quickly and non-destructively, and guide field operations.
MethodHigh-resolution images of rice canopy under bacterial blight stress were acquired using utilized unmanned aerial vehicles (UAVs). Color features and texture features were extracted from the images, and multiple regression models based on color features, texture features, and the fusion of color and texture features were constructed to predict the infection level of rice bacterial blight. The influence of different ground sampling distances (GSD) on the accuracy of the models was also explored.
ResultThe determination coefficient (R2) of the monitoring model based on color features was 85.9%, root mean square error (RMSE) was 1.43 and relative RMSE (RRMSE) was 19.1%. The R2 had increased by 2.4 percentage points and RRMSE had increased by 4.6 percentage points compared with the model based on texture features. Compared with single-feature models, the prediction model based on the fusion of color and texture features (R2=89.6%, RMSE=1.06, RRMSE=15.1%) exhibited significant improvement in accuracy. By constructing models with different GSDs, it was found that when the GSD was 0.2, 0.5 or 0.8 cm, the models achieved higher accuracy with R2 all above 80%.
ConclusionThe color and texture features extracted from low-altitude remote sensing images captured by UAVs can be used for monitoring rice bacterial blight. The results can provide effective technical support for UAV remote sensing monitoring of rice bacterial blight.
-
Keywords:
- Rice /
- Bacterial leaf blight /
- UAV /
- Color feature /
- Texture feature /
- Ground sampling distance (GSD)
-
-
![]()
图 3 基于颜色特征(a)、纹理特征(b)和两者融合(c)的白叶枯感染率预测模型精度检验
Figure 3. Accuracy evaluation of prediction model of bacterial blight infection rate based on color features (a), texture features (b) and their fusion (c)
![]()
图 4 不同GSD条件下白叶枯4个感染等级样本
Figure 4. Samples of four infection levels in rice bacterial blight under different GSD conditions
![]()
图 5 区域尺度水稻白叶枯感染等级分布
Figure 5. Regional scale distribution of rice bacterial blight infection levels
表 1 水稻白叶枯HSV颜色特征参数1)
Table 1 HSV color features parameters of rice bacterial blight
叶片感染等级
Infection level of rice leaves$ {{\boldsymbol{H}}}_{{\boldsymbol{P}}} $ $ {{\boldsymbol{S}}}_{{\boldsymbol{P}}} $ $ {{\boldsymbol{V}}}_{{\boldsymbol{P}}} $ $ {{\boldsymbol{\mu}} }_{{\boldsymbol{H}}} $ $ {{\boldsymbol{\delta}} }_{{\boldsymbol{H}}} $ $ {{\boldsymbol{\mu}} }_{{\boldsymbol{S}}} $ $ {{\boldsymbol{\delta}} }_{{\boldsymbol{S}}} $ $ {{\boldsymbol{\mu}} }_{{\boldsymbol{V}}} $ $ {{\boldsymbol{\delta}} }_{{\boldsymbol{V}}} $ 健康 Healthy 44.33 1110.57 129.16 8109.449 95.41 7645.89 轻微 Slight 41.73 1241.46 139.03 9516.93 83.03 7882.55 中度 Moderate 40.59 1164.32 143.80 9241.22 88.44 8289.44 严重 Serious 39.62 1166.58 148.84 9263.31 89.82 8817.79 1) $ {{\boldsymbol{H}}}_{{\boldsymbol{P}}} $、$ {{\boldsymbol{S}}}_{{\boldsymbol{P}}} $、$ {{\boldsymbol{V}}}_{{\boldsymbol{P}}} $分别表示图像H、S、V颜色分量；$ {\boldsymbol{\mu}} $表示HSV各分量的一阶矩阵；$ {\boldsymbol{\delta}} $表示HSV各分量的二阶矩阵。
1) $ {{\boldsymbol{H}}}_{{\boldsymbol{P}}} $, $ {{\boldsymbol{S}}}_{{\boldsymbol{P}}} $ and $ {{\boldsymbol{V}}}_{{\boldsymbol{P}}} $ represent the color components H, S, and V of the image, respectively; μ represents the first-order moment of each HSV component, and δ represents the second-order moment of each HSV component.
下载: 导出CSV
表 2 不同GSD条件下白叶枯感染等级预测精度
Table 2 Prediction accuracy of bacterial blight infection levels under different GSD conditions
特征
Feature地面分辨率/cm
GSD决定系数/%
R2均方根误差
RMSE相对均方根误差/%
RRMSE颜色
Color0.1 85.9 1.43 19.1 0.2 88.1 1.19 18.6 0.5 85.8 0.86 25.9 0.8 80.9 0.51 28 1.0 74.0 0.43 62.0 纹理
Texture0.1 83.5 1.02 14.5 0.2 84.6 0.93 13.2 0.5 82.4 0.67 19.1 0.8 74.6 0.17 12.1 1.0 68.8 0.15 22.6 颜色和纹理融合
Fusion of color and texture0.1 89.6 1.06 15.1 0.2 91.9 1.15 14.7 0.5 86.6 0.32 10.1 0.8 84.9 0.98 41.7 1.0 73.8 0.12 23.1
下载: 导出CSV
-
[1] RAJARAJESWARI N V L, MURALIDHARAN K. Assessments of farm yield and district production loss from bacterial leaf blight epidemics in rice[J]. Crop Protection, 2006, 25(3): 244-252. doi: 10.1016/j.cropro.2005.04.013
[2] 伍尚忠. 水稻白叶枯病及其防治[M]. 上海: 上海科学技术出版社, 1983. [3] 中国农业科学院植物保护研究所. 中国农作物病虫害(上册)[M]. 北京: 中国农业出版社, 1995. [4] 杨贵军, 李长春, 于海洋, 等. 农用无人机多传感器遥感辅助小麦育种信息获取[J]. 农业工程学报, 2015, 31(21): 184-190. doi: 10.11975/j.issn.1002-6819.2015.21.024 [5] 张智韬, 边江, 韩文霆, 等. 无人机热红外图像计算冠层温度特征数诊断棉花水分胁迫[J]. 农业工程学报, 2018, 34(15): 77-84. doi: 10.11975/j.issn.1002-6819.2018.15.010 [6] 杨文攀, 李长春, 杨浩, 等. 基于无人机热红外与数码影像的玉米冠层温度监测[J]. 农业工程学报, 2018, 34(17): 68-75. doi: 10.11975/j.issn.1002-6819.2018.17.010 [7] DEHKORDI R H, JARROUDI M E, KOUADIO L, et al. Monitoring wheat leaf rust and stripe rust in winter wheat using high-resolution UAV-based red-green-blue imagery[J]. Remote Sensing, 2020, 12(22): 3696. doi: 10.3390/rs12223696
[8] 王震, 褚桂坤, 张宏建, 等. 基于无人机可见光图像Haar-like特征的水稻病害白穂识别[J]. 农业工程学报, 2018, 34(20): 73-82. doi: 10.11975/j.issn.1002-6819.2018.20.010 [9] WEI L L, LUO Y S, XU L Z, et al. Deep convolutional neural network for rice density prescription map at ripening stage using unmanned aerial vehicle-based remotely sensed images[J]. Remote Sensing, 2022, 14(1): 46.
[10] DANG L M, WANG H X, LI Y F, et al. Fusarium wilt of radish detection using RGB and near infrared images from unmanned aerial vehicles[J]. Remote Sensing, 2020, 12(17): 2863. doi: 10.3390/rs12172863
[11] ZHANG D, ZHOU X, ZHANG J, et al. Detection of rice sheath blight using an unmanned aerial system with high-resolution color and multispectral imaging[J]. PLoS One, 2018, 13(5): e0187470. doi: 10.1371/journal.pone.0187470.
[12] LEE K D, KIM S M, AHN H Y, et al. Yearly estimation of rice growth and bacterial leaf blight inoculation effect using UAV imagery[J]. Journal of the Korean Society of Agricultural Engineers, 2020, 62(4): 75-86.
[13] 赵晓阳, 张建, 张东彦, 等. 低空遥感平台下可见光与多光谱传感器在水稻纹枯病病害评估中的效果对比研究[J]. 光谱学与光谱分析, 2019, 39(4): 1192-1198. [14] LIU T, SHI T Z, ZHANG H, et al. Detection of rise damage by leaf folder (Cnaphalocrocis medinalis) Using unmanned aerial vehicle based hyperspectral data[J]. Sustainability, 2020, 12(22): 9343. doi: 10.3390/su12229343
[15] WANG T Y, THOMASSON J A, ISAKEIT T, et al. A plant-by-plant pethod to pdentify and treat cotton root rot based on UAV remote sensing[J]. Remote Sensing, 2020, 12(15): 2453. doi: 10.3390/rs12152453
[16] SU J Y, LIU C J, HU X P, et al. Spatio-temporal monitoring of wheat yellow rust using UAV multispectral imagery[J]. Computers and Electronics in Agriculture, 2019, 167: 105035. doi: 10.1016/j.compag.2019.105035.
[17] 万亮, 岑海燕, 朱姜蓬, 等. 基于纹理特征与植被指数融合的水稻含水量无人机遥感监测[J]. 智慧农业, 2020, 2(1): 58-67. doi: 10.12133/j.smartag.2020.2.1.201911-SA002 [18] WANG C W, CHEN Y C, XIAO Z P, et al. Cotton blight identification with ground framed canopy photo-assisted multispectral UAV images[J]. Agronomy, 2023, 13(5): 1222. doi: 10.3390/agronomy13051222
[19] DAMMER K H, GARZ A, HOBART M, et al. Combined UAV-and tractor-based stripe rust monitoring in winter wheat under field conditions[J]. Agronomy Journal, 2022, 114(1): 651-661. doi: 10.1002/agj2.20916
[20] XAVIER T W F, SOUTO R N V, STATELLA T, et al. Identification of Ramularia leaf blight cotton disease infection levels by multispectral, multiscale UAV imagery[J]. Drones, 2019, 3(2): 33. doi: 10.3390/drones3020033
[21] BHANDARI M, IBRAHIM A M H, XUE Q W, et al. Assessing winter wheat foliage disease severity using aerial imagery acquired from small Unmanned Aerial Vehicle (UAV)[J]. Computers and Electronics in Agriculture, 2020, 176: 105665. doi: 10.1016/j.compag.2020.105665.
[22] ZHANG T X, YANG Z F, XU Z Y, et al. Wheat yellow rust severity detection by efficient DF-UNet and UAV multispectral imagery[J]. IEEE Sensors Journal, 2022, 22(9): 9057-9068. doi: 10.1109/JSEN.2022.3156097
[23] HARSHADKUMAR B P, JITESH P S, VIPUL K D. Detection and classification of rice plant diseases[J]. Intelligent Decision Technologies, 2018, 11(3): 357-373.
[24] XU W J, KESHMIRI S, WANG G. Adversarially approximated autoencoder for image generation and manipulation[J]. IEEE Transactions on Multimedia, 2019, 21(9): 2387-2396. doi: 10.1109/TMM.2019.2898777
[25] GÖRGEL P, SIMSEK A. Face recognition via deep stacked denoising sparse autoencoders (DSDSA)[J]. Applied Mathematics and Computation, 2019, 355: 325-342. doi: 10.1016/j.amc.2019.02.071
