Bistatic measurements for the estimation of rice crop variables using artificial neural network

Gupta, Dileep Kumar; Kumar, Pradeep; Mishra, Varun Narayan; Prasad, Rajendra; Dikshit, Prabhat Kumar Singh; Dwivedi, Shridhar; Ohri, Anurag; Singh, Ravi Shankar; Srivastava, V. K.

doi:10.1016/j.asr.2015.01.003

Cited by 26 publications

(10 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Artificial neural networks (ANNs) are a commonly used tool in remote sensing of the Earth's surface that often achieves the same accuracies as RF and SVM (Petropoulos, Kontoes, & Keramitsoglou, 2012). They have been successfully applied to estimation of biophysical forest parameters for multispectral data (Linderman et al, 2004), biomass and soil moisture retrieval (Ali, Greifeneder, Stamenkovic, Neumann, & Notarnicola, 2015), crop monitoring (Gupta et al, 2015) or land cover classification (Petropoulos et al, 2012). Despite all those uses for ANN, the topic of tree species classification using ANN is seldom approached by the research community.…”

Section: Introductionmentioning

confidence: 99%

Comparison of support vector machine, random forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images

Raczko

Zagajewski

2017

European Journal of Remote Sensing

293

176

View full text Add to dashboard Cite

Knowledge of tree species composition in a forest is an important topic in forest management. Accurate tree species maps allow for much more detailed and in-depth analysis of biophysical forest variables. The paper presents a comparison of three classification algorithms: support vector machines (SVM), random forest (RF) and artificial neural networks (ANN) for tree species classification using airborne hyperspectral data from the Airborne Prism EXperiment sensor. The aim of this paper is to evaluate the three nonparametric classification algorithms (SVM, RF and ANN) in an attempt to classify the five most common tree species of the Szklarska Poręba area: spruce (Picea alba L. Karst), larch (Larix decidua Mill.), alder (Alnus Mill), beech (Fagus sylvatica L.) and birch (Betula pendula Roth). To avoid human introduced biases a 0.632 bootstrap procedure was used during evaluation of each compared classifier. Of all compared classification results, ANN achieved the highest median overall classification accuracy (77%) followed by SVM with 68% and RF with 62%. Analysis of the stability of results concluded that RF and SVM had the lowest variance of overall accuracy and kappa coefficient (12 percentage points) while ANN had 15 percentage points variance in results.

show abstract

Section: Introductionmentioning

confidence: 99%

Comparison of support vector machine, random forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images

Raczko

Zagajewski

2017

European Journal of Remote Sensing

293

176

View full text Add to dashboard Cite

show abstract

“…Several researchers have developed robust machine learning algorithms by selecting the appropriate model parameters using error-trial methods, which provided the encouraging results. [18][19][20] Frate et al 21 have trained two separate artificial neural networks algorithms by physical vegetation model to retrieve the soil moisture and wheat crop variables using L-band, dual-polarized, and multiangular radiometric data. They found the retrieval process quite effective.…”

Section: Introductionmentioning

confidence: 99%

Robust machine learning techniques for rice crop variables estimation using multiangular bistatic scattering coefficients

Gupta

Prasad

Kumar

et al. 2018

J. Appl. Rem. Sens.

View full text Add to dashboard Cite

The present study is designed to explore the potential of bistatic scattering coefficients (σ°) and machine learning algorithms for the estimation of rice crop variables using groundbased multiangular, multitemporal, and dual-polarized bistatic scatterometer data. The bistatic scatterometer measurements are carried out at eight different growth stages of the rice crop in the angular range of incidence angle 20 deg to 70 deg for HH-and VV-polarization at 10-GHz frequency in the specular direction with an azimuthal angle (φ ¼ 0). Several field measurements are taken for the measurement of rice crop variables, such as vegetation water content, leaf area index, and plant height at its various growth stages. Machine learning algorithms-such as fuzzy inference system (FIS), support vector machine for regression (SVR), and generalized linear model (GLM)-are used to estimate the rice crop variables using bistatic scatterometer data. The linear regression analysis is carried out for the evaluation of the multiangular, multitemporal, and dual-polarized datasets for the selection of optimum incidence angle and polarization for accurate estimation of rice crop variables. The highest value of the coefficient of determination (R 2) is found at 30-deg incidence angle for VV-polarization. The sensitivity of copolarized ratio of σ°with the rice crop variable is also evaluated using linear regression analysis for the estimation of rice crop variables. The highest value of R 2 is found to be at 35-deg incidence angle between the copolarized ratio of σ°and rice crop variables. The performance of SVR model is found superior in comparison to the FIS and GLM at VV-polarization and the copolarized ratio of σ°for the estimation of rice crop variables. However, the copolarized ratio of σ°is found superior to VV-polarized bistatic scatterometer data for the estimation of rice crop variables.

show abstract

“…Meanwhile, electromagnetic scattering and propagation in vegetations play important roles in certain fields such as agricultural production [1], environment change prediction [2], wireless communications [3] and target detection [4]. Recently, research on bistatic scattering [5][6][7][8][9][10] has received considerable attention because of the multidimensional information provided by bistatic radar systems, which is more than that provided by the monostatic configuration for it only contains 1D data in the backscattering direction. Additional information contained in the bistatic radar has the potential to provide improvements in monitoring vegetation growth and detecting drought degree of crops.…”

Section: Introductionmentioning

confidence: 99%

“…In 2000, Ferrazzoli et al [27] found that the bistatic scattering configuration can overcome the saturation problems occurring in vegetation biomass retrieval at the backscattering direction in L and C bands. In 2015, Gupta et al [8] measured the rice bistatic scattering coefficients in temporal changes at X band and retrieved crop growth variables using two types of neural network models. When it comes to the work of bistatic scattering interactions between incident waves and wheat and soybean parameters, few research works have been conducted.…”

Section: Introductionmentioning

confidence: 99%

Bistatic Scattering Characteristics of Wheat and Soybean by Radiative Transfer Model in L Band and C Band

Zhang¹,

Wu²

2016

PIER B

View full text Add to dashboard Cite

Compared with the backscattering configuration, the bistatic scattering echoes can provide multidimensional information on land surface. Based on the Michigan Microwave Canopy Scattering (MIMICS) model, a first-order microwave bistatic scattering model for vegetations is developed in this paper. The dominant scattering mechanism for wheat and soybean in the L and C bands is analyzed by simulating the bistatic scattering echoes in multiple viewpoints, which can help us understand the interaction between incident wave and vegetation parameters. The influence of crop height, leaf size and moisture of vegetations and down layer soil on the scattering echoes is fully investigated. The simulations show that the bistatic scattering echoes are more sensitive to the vegetation parameters than that in backscattering configuration. There exist optimal scattering angles, in specular direction and in direction perpendicular to the incident plane, to improve the retrieval accuracy of vegetation parameters and moisture of soil surface. Moreover, the simulations demonstrate that bistatic scattering echoes in high frequency (C band) are a good choice to retrieve the vegetation parameters, and the echoes in low frequency (L band) are preferred to retrieve the soil parameters. This research can be used to provide reference for crop monitoring and future bistatic system design.

show abstract

Bistatic measurements for the estimation of rice crop variables using artificial neural network

Cited by 26 publications

References 34 publications

Comparison of support vector machine, random forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images

Comparison of support vector machine, random forest and neural network classifiers for tree species classification on airborne hyperspectral APEX images

Robust machine learning techniques for rice crop variables estimation using multiangular bistatic scattering coefficients

Bistatic Scattering Characteristics of Wheat and Soybean by Radiative Transfer Model in L Band and C Band

Contact Info

Product

Resources

About