[Show abstract][Hide abstract] ABSTRACT: Powdery mildew is one of the most serious diseases that have a significant impact on the production of winter wheat. As an effective alternative to traditional sampling methods, remote sensing can be a useful tool in disease detection. This study attempted to use multi-temporal moderate resolution satellite-based data of surface reflectances in blue (B), green (G), red (R) and near infrared (NIR) bands from HJ-CCD (CCD sensor on Huanjing satellite) to monitor disease at a regional scale. In a suburban area in Beijing, China, an extensive field campaign for disease intensity survey was conducted at key growth stages of winter wheat in 2010. Meanwhile, corresponding time series of HJ-CCD images were acquired over the study area. In this study, a number of single-stage and multi-stage spectral features, which were sensitive to powdery mildew, were selected by using an independent t-test. With the selected spectral features, four advanced methods: mahalanobis distance, maximum likelihood classifier, partial least square regression and mixture tuned matched filtering were tested and evaluated for their performances in disease mapping. The experimental results showed that all four algorithms could generate disease maps with a generally correct distribution pattern of powdery mildew at the grain filling stage (Zadoks 72). However, by comparing these disease maps with ground survey data (validation samples), all of the four algorithms also produced a variable degree of error in estimating the disease occurrence and severity. Further, we found that the integration of MTMF and PLSR algorithms could result in a significant accuracy improvement of identifying and determining the disease intensity (overall accuracy of 72% increased to 78% and kappa coefficient of 0.49 increased to 0.59). The experimental results also demonstrated that the multi-temporal satellite images have a great potential in crop diseases mapping at a regional scale.
PLoS ONE 01/2014; 9(4):e93107. · 3.73 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Detection of yellow rust is of great importance in disease control and reducing the use of fungicide. Spectral analysis is an important method for disease detection in terms of remote sensing. In this study, an emerging spectral analysis method known as continuous wavelet analysis (CWA) was examined and compared with several conventional spectral features for the detection of yellow rust disease at a leaf level. The leaf spectral measurements were made by a spectroradiometer at both Zodaks 37 and 70 stages with a large sample size. The results showed that the wavelet features were able to capture the major spectral signatures of yellow rust, and exhibited considerable potential for disease detection at both growth stages. Both the accuracies of the univariate and multivariate models suggested that wavelet features outperformed conventional spectral features in quantifying disease severity at leaf level. Optimal accuracies returned a coefficient of determination (R2) of 0.81 and a root mean square error (RMSE) of 0.110 for pooled data at both stages. Furthermore, wavelet features showed a stronger response to the yellow rust at Zodaks 70 stage than at Zodaks 37 stage, indicating reliable estimation of disease severity can be made until the Zodaks 70 stage.
Computers and Electronics in Agriculture 01/2014; 100:79–87. · 1.77 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Yellow rust (Puccinia striiformis f. sp. Tritici), powdery mildew (Blumeria graminis) and wheat aphid (Sitobion avenae F.) infestation are three serious conditions that have a severe impact on yield and grain quality of winter wheat worldwide. Discrimination among these three stressors is of practical importance, given that specific procedures (i.e. adoption of fungicide and insecticide) are needed to treat different diseases and insects. This study examines the potential of hyperspectral sensor systems in discriminating these three stressors at leaf level. Reflectance spectra of leaves infected with yellow rust, powdery mildew and aphids were measured at the early grain filling stage. Normalization was performed prior to spectral analysis on all three groups of samples for removing differences in the spectral baseline among different cultivars. To obtain appropriate bands and spectral features (SFs) for stressor discrimination and damage intensity estimation, a correlation analysis and an independent t-test were used jointly. Based on the most efficient bands/SFs, models for discriminating stressors and estimating stressor intensity were established by Fisher's linear discriminant analysis (FLDA) and partial least square regression (PLSR), respectively. The results showed that the performance of the discrimination model was satisfactory in general, with an overall accuracy of 0.75. However, the discrimination model produced varied classification accuracies among different types of diseases and insects. The regression model produced reasonable estimates of stress intensity, with an R2 of 0.73 and a RMSE of 0.148. This study illustrates the potential use of hyperspectral information in discriminating yellow rust, powdery mildew and wheat aphid infestation in winter wheat. In practice, it is important to extend the discriminative analysis from leaf level to canopy level.
Field Crops Research 01/2014; 156:199–207. · 2.47 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Crop aboveground biomass estimates are critical for assessing crop growth and predicting yield. In order to ascertain the optimal methods for winter wheat biomass estimation, this study compared the utility of univariate techniques involving narrow band vegetation indices and red-edge position (REP), as well as multivariate calibration techniques involving the partial least square regression (PLSR) analyses using band depth parameters, and the combination of band depth parameters and hyperspectral indices including narrow band indices and REP. Narrow band indices were calculated in the form of normalized difference vegetation index (NDVI) and soil adjusted vegetation index (SAVI) using all possible two-band combinations for selecting optimal narrow band indices. Band depth, band depth ratio (BDR), normalized band depth index, and band depth normalized to area extracted from a red absorption region (550 nm–750 nm) were utilized as band depth parameters. The results indicated that: (1) Compared with the traditional NDVI and SAVI constructed with bands at 670 nm and 800 nm and REP, the selected narrow band indices (optimal NDVI-like and optimal SAVI-like) produced higher estimation accuracy of the winter wheat biomass; (2) the PLSR models based on band depth parameters produced lower root mean square error, relative to the models based on the selected narrow band indices; and (3) the PLSR model based on the combination of optimal NDVI-like and BDR produced the best estimated result of the winter wheat biomass (R2 = 0.84, RMSE = 0.177 kg/m2). The results of this study suggest that PLSR analysis using the combination of optimal NDVI-like and band depth parameters could significantly improve estimation accuracy of winter wheat biomass.
Computers and Electronics in Agriculture 01/2014; 100:51–59. · 1.77 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: A fast and reliable ultra-high performance liquid chromatography–tandem mass spectrometry method was developed for the determination of aflatoxins B1, B2, G1, and G2 in cereal. The analytes were extracted by accelerated solvent extraction with methanol/water (80:20). A polymeric solid-phase extraction column was used for sample preparation. Under optimum conditions, the analyte recoveries for samples spiked at different concentration levels in rice and maize ranged from 71.2 to 94.0%, with relative standard deviations less than 16.4%. Limits of detection (signal-to-noise ratio, 3:1) for the aflatoxins ranged from 0.25 to 0.93 ng/g. The developed method was applied to the determination of aflatoxins in ten rice and maize samples. One maize sample tested positive with an aflatoxin B1 concentration of 2.7 ng/g.
[Show abstract][Hide abstract] ABSTRACT: The objective of this study was to estimate soil moisture with soil red and nir (near-infrared) band reflectance from TM/ETM+ remotely sensed images acquired over vegetated fields. Based on linear decomposition algorithm of mixture pixel, first the soil reflectance from red–nir bands were directly and computationally derived by combining soil line equation with a developed empirical relationship between vegetation canopy and mixture pixel reflectance in red–nir spectral feature space. Then, a remote sensing image from TM with measurement data from experimental fields in Beijing, China, was used to establish the retrieval relationships between soil moisture and soil reflectance from the red and nir bands, and the results showed that the retrieval of soil moisture was better with nir band reflectance than that of red. Finally, the soil moisture retrieved method was further evaluated and validated with two images from ETM + and ground measurements from fields in Walnut Creek, America, and the analysis showed that the proposed method could be used to monitor soil moisture well, with the correlation coefficient exceeding 0.80. The preliminary results with such acceptable accuracy indicate that the method of estimating soil moisture based on the linear decomposition of mixture pixels is reasonable and suitable for being widely applied in different temporal and spatial scaled fields.
Mathematical and Computer Modelling. 08/2013; 58(s 3–4):606–613.
[Show abstract][Hide abstract] ABSTRACT: Leaf area index (LAI) is a major indicator for crop growth monitoring and yield estimation. Data assimilation as an effective tool for crop LAI estimation fully considers the properties of actual observations and physical model simulations. In this work, we present a new data assimilation scheme, introducing a very fast simulated annealing (VFSA) optimization algorithm into the process of crop LAI assimilation with a four dimensional variational data assimilation (4DVAR) algorithm. Firstly, calibrating the input parameters of a crop growth simulation model based on history observations. Secondly, quantitatively describing the relationship between fields observed data and model simulated data by the cost function of 4DVAR algorithm. Finally, the optimization process of the cost function is accomplished by the VFSA optimization algorithm, and further the optimal solution is taken as the best combination of input parameters of physical model for LAI estimation. Winter wheat in Beijing is taken as an experimental object. The numerical results show not only the improved time efficiency of this proposed assimilation scheme, but also enhanced assimilation accuracy of all LAI assimilations, especially for LAI ≥≥ 3.00. Theoretical analysis and practical experiments confirm the application prospect of VFSA optimization algorithm in LAI variational assimilation.
[Show abstract][Hide abstract] ABSTRACT: Since van Groenigen and Stein (1998) proposed the SSA+MMSD (Spatial Simulated Annealing + Minimization of the Mean of Shortest Distances criterion) method, this method has received wide application in the optimization of sampling designs. However, it is computationally inefficient due to the complexity of this method itself. Initially in this paper, we analyze the computational complexity associated with this method from both SSA and MMSD aspects. And then, we propose some corresponding revisions (including the initial solution, perturbation rules, as well as the objective function) accordingly so as to reduce its computations. Finally, we evaluate the efficiency improvement via comparing some efficiency indexes of both original and modified methods (including the total perturbations needed, valid and better candidate designs generating rates of the perturbations, and the rate of objective function decline). Analysis and experimental results indicate that the modified method is much more efficient than the original one; in C++ implementations, the mean execution time needed for the modified method is only about 1/3 of that of the original.
Mathematical and Computer Modelling - MATH COMPUT MODELLING. 08/2013;
[Show abstract][Hide abstract] ABSTRACT: or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. V egetation water content (VWC) is an impor-tant biophysical variable that is often used in agricultural and ecological remote sensing applications (Ceccato et al., 2002a, 2002b; Gao and Goetz, 1995; Ustin et al., 1998). The concept of VWC is complicated by the largely unanswered question of precisely what event or condition defines the onset of water deficit stress in plants (Jackson and Ezra, 1985; Jackson et al., 1986). The measurement of VWC is important because it can be calculated from remotely sensed data col-lected by current satellites (Allen et al., 1971; Carter, 1994; Clevers et al., 2008, 2010; Claudio et al., 2006). Tucker (1980) and Gausman (1985) found that the shortwave near-infrared wavebands (1400–2500 nm), in which water absorption features are observed, were heavily influenced by water in plant tissues. Specifically, the wavelengths at 1530 and 1720 nm appeared to be most sensitive to VWC changes, and therefore these bands may be seen as appropriate for the assessment of VWC. Penuelas et al. (1993) defined the so-called water band index, which is the ratio between reflectance at 970 nm and reflectance at 900 nm. Gao (1996) defined the normalized difference water index (NDWI), in which the 1200-and 860-nm water absorption feature wavelengths were used. In the estimation of VWC, hyperspectral reflectance data, represent-ing a range of canopies, were simulated using the combined PROSPECT + SAILH model and spectral derivatives at the slopes of the 970-and 1200-nm water absorption features (Clevers et al., 2008, 2010). Wang et al. (2011b) reported that the normalized difference matter index (NDMI) could estimate the canopy foliar biomass using spectral reflectance measurements. Claudio et al. (2006) monitored drought effects on the VWC and fluxes in chaparral using the 970-nm water band index. Colombo et al. (2008) showed that the estima-tion precision of the leaf equivalent water thickness (EWT) could be improved by the use of inverse ordinary least squares regression and reduced major axis regression; however, the inversion of the PROSPECT model revealed some challenges inherent in the simultaneous estimation of leaf EWT. Seelig et al. (2008) demonstrated that the T 1300 /T 1450 leaf water index, which is based on leaf light transmission, demonstrated a strong exponential correlation with leaf relative water thickness (LRWT). This was revealed first through theoretical analysis and was then applied for effective estimation of cowpea [Vigna unguiculata (L.) Walp.], bean (Vicia faba L.), and sugarbeet (Beta vulgaris L.) LRWT. Sancho-Knapik et al. (2011) used the microwave l-band (1730 MHz) to accurately estimate relative water content in poplar leaves (Populus trichocarpa Torr. & A. Gra P. deltoides W. Bartram ex Marshall). The three-band algorithm was designed to combine three different sensitivity bands, such as chlorophyll and water
[Show abstract][Hide abstract] ABSTRACT: Abstract—The nitrogen nutrition index (NNI) is calculated
from the measured N concentration and the critical nitrogen (N)
curve. It can be used to determine the N required by a crop and
is helpful for optimizing N application in the field. Our objectives
were to validate the existing corn critical N curve for the
northwestern plain of Shandong Province and to design a more
accurate remote detection method for the NNI. For this purpose,
field measurements were conducted weekly to acquire the biomass
and N concentrations during the corn growing season of 2011.
Additionally, nearly 60 corn canopy spectra were collected during
field campaigns. First, limiting and non-limiting N points were
selected from sampled data, and they were used to validate the
existing critical N curve. Second, an NNI estimation model based
on a Principal Component Analysismethod and Back Propagation
Artificial Neural Network (PCA-BP-ANN) model was established.
The collected canopy spectra and corresponding NNI were used
to compare the performances of the above mentioned method and
other for NNI estimation. The results showed that the N curve
proposed in the literature is suitable for the study region. Among
the three remote detection methods, PCA-BP-ANN provided the
best results with highest R value and lowest root mean square
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 04/2013; 6(2):682-689. · 2.87 Impact Factor