Enhanced Regional Monitoring of Wheat Powdery Mildew Based on an Instance-Based Transfer Learning Method

In order to monitor the prevalence of wheat powdery mildew, current methods require sufficient sample data to obtain results with higher accuracy and stable validation. However, it is difficult to collect data on wheat powdery mildew in some regions, and this limitation in sampling restricts the accuracy of monitoring regional prevalence of the disease. In this study, an instance-based transfer learning method, i.e., TrAdaBoost, was applied to improve the monitoring accuracy with limited field samples by using auxiliary samples from another region. By taking into account the representativeness of contributions of auxiliary samples to adjust the weight placed on auxiliary samples, an optimized TrAdaBoost algorithm, named OpTrAdaBoost, was generated to map regional wheat powdery mildew. The algorithm conducts this by: (1) producing uncertainty associated with each prediction based on the similarities, and calculating the representativeness contribution of all auxiliary samples by taking into account the overall uncertainty of the wheat powdery mildew map; (2) calculating the errors of the weak learners during the training process and using boosting to filter out the unreliable auxiliary samples by adjusting the weights of auxiliary samples; (3) combining all weak learners according to the weights of training instances to build a strong learner to classify disease severity. OpTrAdaBoost was tested using a dataset with 39 study area samples and 106 auxiliary samples. The overall monitoring accuracy was 82%, and the kappa coefficient was 0.72. Moreover, OpTrAdaBoost performed better than other algorithms that are commonly used to monitor wheat powdery mildew at the regional level. Experimental results demonstrated that OpTrAdaBoost was effective in improving the accuracy of monitoring wheat powdery mildew using limited field samples.

[1]  J. Liu,et al.  Predictive soil mapping with limited sample data , 2015 .

[2]  Yong Luo,et al.  Detection of powdery mildew in two winter wheat cultivars using canopy hyperspectral reflectance , 2013 .

[3]  Naif Alajlan,et al.  Using convolutional features and a sparse autoencoder for land-use scene classification , 2016 .

[4]  Min Zhang,et al.  Temporal and spatial dynamics of wheat powdery mildew in Sichuan Province, China , 2015 .

[5]  Ronghua Ma,et al.  Predicting the probability of wheat aphid occurrence using satellite remote sensing and meteorological data , 2014 .

[6]  N. M. Faber,et al.  How to avoid over-fitting in multivariate calibration--the conventional validation approach and an alternative. , 2007, Analytica chimica acta.

[7]  Stefano Ermon,et al.  Monitoring Ethiopian Wheat Fungus with Satellite Imagery and Deep Feature Learning , 2017, 2017 IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).

[8]  Qiang Yang,et al.  Boosting for transfer learning , 2007, ICML '07.

[9]  Cunfu Liu,et al.  Application of hydrochemistry and stable isotopes (δ34S, δ18O and δ37Cl) to trace natural and anthropogenic influences on the quality of groundwater in the piedmont region, Shijiazhuang, China , 2016 .

[10]  Lin Yuan,et al.  Comparison of Methods for Forecasting Yellow Rust in Winter Wheat at Regional Scale , 2014, CCTA.

[11]  Hannes Taubenböck,et al.  Class imbalance in unsupervised change detection - A diagnostic analysis from urban remote sensing , 2017, Int. J. Appl. Earth Obs. Geoinformation.

[12]  Boleslo E. Romero,et al.  A quasi-global precipitation time series for drought monitoring , 2014 .

[13]  Yi Guan,et al.  Transfer learning based clinical concept extraction on data from multiple sources , 2014, J. Biomed. Informatics.

[14]  Baofeng Su,et al.  Significant Remote Sensing Vegetation Indices: A Review of Developments and Applications , 2017, J. Sensors.

[15]  Bernhard E. Boser,et al.  A training algorithm for optimal margin classifiers , 1992, COLT '92.

[16]  Hui Qing Liu,et al.  A feedback based modification of the NDVI to minimize canopy background and atmospheric noise , 1995, IEEE Transactions on Geoscience and Remote Sensing.

[17]  Jihua Wang,et al.  Damage Mapping of Powdery Mildew in Winter Wheat with High-Resolution Satellite Image , 2014, Remote. Sens..

[18]  B. Markham,et al.  Summary of Current Radiometric Calibration Coefficients for Landsat MSS, TM, ETM+, and EO-1 ALI Sensors , 2009 .

[19]  Xi Zhou,et al.  Data augmentation for face recognition , 2017, Neurocomputing.

[20]  Wei Gong,et al.  Transfer learning used to analyze the dynamic evolution of the dust aerosol , 2015 .

[21]  Fawaz S. Al-Anzi,et al.  Employing fisher discriminant analysis for Arabic text classification , 2017, Comput. Electr. Eng..

[22]  Wei Liu,et al.  Regionalization of wheat powdery mildew oversummering in China based on digital elevation , 2018 .

[23]  Wenjiang Huang,et al.  Monitoring Powdery Mildew of Winter Wheat by Using Moderate Resolution Multi-Temporal Satellite Imagery , 2014, PloS one.

[24]  Lorenzo Bruzzone,et al.  Domain Adaptation for the Classification of Remote Sensing Data: An Overview of Recent Advances , 2016, IEEE Geoscience and Remote Sensing Magazine.

[25]  Georgios Tziritas,et al.  MRF-based segmentation and unsupervised classification for building and road detection in peri-urban areas of high-resolution satellite images , 2016 .

[26]  Derek T. Anderson,et al.  Comprehensive survey of deep learning in remote sensing: theories, tools, and challenges for the community , 2017 .

[27]  Jiangang Fei,et al.  Use of exploratory factor analysis in maritime research , 2018, The Asian Journal of Shipping and Logistics.

[28]  H. J. Li,et al.  Identification and genetic mapping of the putative Thinopyrum intermedium-derived dominant powdery mildew resistance gene PmL962 on wheat chromosome arm 2BS , 2015, Theoretical and Applied Genetics.

[29]  William J. Emery,et al.  The Importance of Physical Quantities for the Analysis of Multitemporal and Multiangular Optical Very High Spatial Resolution Images , 2014, IEEE Transactions on Geoscience and Remote Sensing.

[30]  Andreas Christmann,et al.  Support vector machines , 2008, Data Mining and Knowledge Discovery Handbook.

[31]  A. Zhao,et al.  Effect of straw management on carbon sequestration and grain production in a maize–wheat cropping system in Anthrosol of the Guanzhong Plain , 2016 .

[32]  Liping Di,et al.  Comparison between TVDI and CWSI for drought monitoring in the Guanzhong Plain, China , 2017 .

[33]  Theofanis Sapatinas,et al.  Discriminant Analysis and Statistical Pattern Recognition , 2005 .

[34]  Dong Liang,et al.  New Optimized Spectral Indices for Identifying and Monitoring Winter Wheat Diseases , 2014, IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing.

[35]  Mostafa Mehdipour-Ghazi,et al.  Plant identification using deep neural networks via optimization of transfer learning parameters , 2017, Neurocomputing.

[36]  Onisimo Mutanga,et al.  Remote sensing of crop health for food security in Africa: Potentials and constraints , 2017 .

[37]  Budiman Minasny,et al.  On digital soil mapping , 2003 .

[38]  Meixue Yang,et al.  Using the NDVI to identify variations in, and responses of, vegetation to climate change on the Tibetan Plateau from 1982 to 2012 , 2017 .

[39]  A. Huete,et al.  Overview of the radiometric and biophysical performance of the MODIS vegetation indices , 2002 .

[40]  Wenjiang Huang,et al.  Evaluation of wavelet spectral features in pathological detection and discrimination of yellow rust and powdery mildew in winter wheat with hyperspectral reflectance data , 2017 .

[41]  Onisimo Mutanga,et al.  Separability of coffee leaf rust infection levels with machine learning methods at Sentinel-2 MSI spectral resolutions , 2017, Precision Agriculture.

[42]  Jeff Dozier,et al.  A generalized split-window algorithm for retrieving land-surface temperature from space , 1996, IEEE Trans. Geosci. Remote. Sens..

[43]  Lin Yuan,et al.  Habitat monitoring to evaluate crop disease and pest distributions based on multi-source satellite remote sensing imagery , 2017 .

[44]  Shuai Wang,et al.  Classifying Wheat Hyperspectral Pixels of Healthy Heads and Fusarium Head Blight Disease Using a Deep Neural Network in the Wild Field , 2018, Remote. Sens..

[45]  Francisco Herrera,et al.  Exploring the effectiveness of dynamic ensemble selection in the one-versus-one scheme , 2017, Knowl. Based Syst..

[46]  Wenjiang Huang,et al.  Detection and discrimination of pests and diseases in winter wheat based on spectral indices and kernel discriminant analysis , 2017, Comput. Electron. Agric..

[47]  Wenjiang Huang,et al.  Integrating Early Growth Information to Monitor Winter Wheat Powdery Mildew Using Multi-Temporal Landsat-8 Imagery , 2018, Sensors.

[48]  John A. Richards,et al.  Remote Sensing Digital Image Analysis , 1986 .

[49]  Qiang Yang,et al.  A Survey on Transfer Learning , 2010, IEEE Transactions on Knowledge and Data Engineering.

[50]  D. Wooff.  Logistic Regression: a Self-learning Text, 2nd edn , 2004 .

[51]  B. K. Bhattacharya,et al.  A multi-stage tracking for mustard rot disease combining surface meteorology and satellite remote sensing , 2013 .

[52]  Na Li,et al.  A transfer learning method for automatic identification of sandstone microscopic images , 2017, Comput. Geosci..

[53]  Bernhard Schölkopf,et al.  Comparing support vector machines with Gaussian kernels to radial basis function classifiers , 1997, IEEE Trans. Signal Process..

[54]  A. Huete,et al.  Optical-Biophysical Relationships of Vegetation Spectra without Background Contamination , 2000 .

[55]  Wenjiang Huang,et al.  Discriminating Wheat Aphid Damage Degree Using 2-Dimensional Feature Space Derived from Landsat 5 TM , 2012 .

[56]  Elaine B. Martin,et al.  Model selection for partial least squares regression , 2002 .

[57]  Fusuo Zhang,et al.  Effects of potassium fertilization on winter wheat under different production practices in the North China Plain , 2013 .

[58]  Xiangming Xu,et al.  Application of Geographic Information Systems to Identify the Oversummering Regions of Blumeria graminis f. sp. tritici in China. , 2013, Plant disease.

[59]  Ruiliang Pu,et al.  Using high spatial resolution satellite imagery for mapping powdery mildew at a regional scale , 2015, Precision Agriculture.

[60]  A. Giebel,et al.  Early detection of Fusarium infection in wheat using hyper-spectral imaging , 2011 .