Dileep Kumar Gupta

Factors affecting fuel rate in Corex process

Abstract The Corex process has been developed as an alternative to the blast furnace where 80–85% non-coking coal and 15–20% coke is used as fuel for heat generation, production of reduction gases and to maintain adequate char bed permeability in the melter–gasifier. Non-coking coals, which can be used in Corex, have to meet certain physical, chemical and high temperature properties for stable process and to attain high performance levels. JSW Steel operates largest Corex based integrated steel plant with two modules each of 0·8 Mtpa capacity where several coals have been used so far and the type of coal used significantly influenced operation. Statistical analysis shows that the significant parameters affecting fuel rate are moisture, volatile matter, slag rate and melting rate. It was observed that at high rate of production, stability and permeability of char bed becomes critical hence coals producing char of high strength after reaction are required.

Comparison of support vector machine, artificial neural network, and spectral angle mapper algorithms for crop classification using LISS IV data

The Resourcesat-2 is a highly suitable satellite for crop classification studies with its improved features and capabilities. Data from one of its sensors, the linear imaging and self-scanning (LISS IV), which has a spatial resolution of 5.8 m, was used to compare the relative accuracies achieved by support vector machine (SVM), artificial neural network (ANN), and spectral angle mapper (SAM) algorithms for the classification of various crops and non-crop covering a part of Varanasi district, Uttar Pradesh, India. The separability analysis was performed using a transformed divergence (TD) method between categories to assess the quality of training samples. The outcome of the present study indicates better performance of SVM and ANN algorithms in comparison to SAM for the classification using LISS IV sensor data. The overall accuracies obtained by SVM and ANN were 93.45% and 92.32%, respectively, whereas the lower accuracy of 74.99% was achieved using the SAM algorithm through error matrix analysis. Results derived from SVM, ANN, and SAM classification algorithms were validated with the ground truth information acquired by the field visit on the same day of satellite data acquisition.

A statistical significance of differences in classification accuracy of crop types using different classification algorithms

Abstract Crop classification is needed to understand the physiological and climatic requirement of different crops. Kernel-based support vector machines, maximum likelihood and normalised difference vegetation index classification schemes are attempted to evaluate their performances towards crop classification. The linear imaging self-scanning (LISS-IV) multi-spectral sensor data was evaluated for the classification of crop types such as barley, wheat, lentil, mustard, pigeon pea, linseed, corn, pea, sugarcane and other crops and non-crop such as water, sand, built up, fallow land, sparse vegetation and dense vegetation. To determine the spectral separability among crop types, the M-statistic and Jeffries–Matusita (J–M) distance methods have been utilised. The results were statistically analysed and compared using Z-test and χ2-test. Statistical analysis showed that the accuracy results using SVMs with polynomial of degrees 5 and 6 were not significantly different and found better than the other classification algorithms.

Dual-polarimetric C-band SAR data for land use/land cover classification by incorporating textural information

The work presented here showed a comprehensive evaluation of dual-polarimetric RISAT-1 data for land use/land cover (LULC) classification. The textural images were extracted with the help of gray-level co-occurrence matrix approach. Analysis of inter-class separability using transformed divergence method was performed to recognize the potential textural images. The best combination of textural images was also identified on the basis of standard deviation of preferred textural images and correlation coefficients. The maximum likelihood classifier-based classification results for different scenarios were compared. Furthermore, various classification algorithms, maximum likelihood classifier (MLC), artificial neural network (ANN), random forest (RF) and support vector machine (SVM), were performed on the best identified scenario in order to observe the most suitable algorithm for LULC classification. The combination of radiometric and their related textural images was found improving the overall classification accuracy than individual datasets. The highest overall classification accuracy was found using SVM (88.97%) followed by RF (88.45%), ANN (83.65%) and MLC (78.18%).

Artificial neural network for crop classification using C-band RISAT-1 satellite datasets

The artificial neural network algorithm has been used for the classification of rice, corn, pigeon pea, green gram, other crop and non-crop classes in Varanasi District, Uttar Pradesh, India. The C-band, dual polarimetric, temporal satellite datasets of RISAT-1 have been carried out in the present study. The separability analysis using transformed divergence and Jefferies Matusita distance methods were compared. The transformed divergence method has shown better separation between the classes in comparison to Jefferies Matusita distance method. The classification results were ground validated. The overall achieved accuracies were 74.21 and 77.36% for satellite data acquired on 9 August 2013 and 28 September 2013 respectively. Results have shown the better classification accuracy using 28th September 2013 data because of almost crops were in the reproductive stage and high reflectivity of the crops at this stage.

Estimation of crop variables using bistatic scatterometer data and artificial neural network trained by empirical models

Multi-temporal and multi-angular bistatic scatterometer measurements were carried out on two similar specially prepared kidney bean crop beds at two frequencies (6GHz and 10GHz) for like polarizations (HH- and VV-). The present study describes the estimation of crop variables and crop covered soil moisture of kidney bean crop using artificial neural network (ANN). The suitable configurations of bistatic scatterometer system were found at 10GHz, 50? incidence angle for the estimation of kidney bean crop variables and 6GHz, 20? incidence angle for the estimation of crop covered soil moisture at VV-polarization by linear regression analysis. Two artificial neural network models namely ANN-I and ANN-II were developed for the estimation of crop variables and crop covered soil moisture of kidney bean crop, respectively. The observed data set (scattering coefficients, crop variables and crop covered soil moisture) of first crop bed of kidney bean was used as a reference data set for developing empirical models. The training of the ANN-I model was done using 95 data set generated through empirical models consistent with the age of the kidney bean crop. The ANN-II was trained using the scattering coefficients and crop covered soil moisture of reference crop bed. The trained ANN-I and ANN-II models were tested by the observed data set of second kidney bean crop bed. The estimated values by ANN-I and ANN-II were found very close to the observed values of the crop variables and crop covered soil moisture of second kidney bean crop bed.

Comprehensive evaluation of soil moisture retrieval models under different crop cover types using C-band synthetic aperture radar data

Abstract In the present study, random forest regression (RFR), support vector regression (SVR) and artificial neural network regression (ANNR) models were evaluated for the retrieval of soil moisture covered by winter wheat, barley and corn crops. SVR with radial basis function kernel was provided the highest adj. R2 (0.95) value for soil moisture retrieval covered by the wheat crop at VV polarization. However, RFR provided the adj. R2 (0.94) value for soil moisture retrieval covered by barley crop at VV polarization using Sentinel-1A satellite data. The adj. R2 (0.94) values were found for the soil moisture covered by corn crop at VV polarization using RFR, SVR linear and radial basis function kernels. The least performance was reported using ANNR model for almost all the crops under investigation. The soil moisture retrieval outcomes were found better at VV polarization in comparison to VH polarization using three different models.

Estimation of winter wheat crop growth parameters using time series Sentinel-1A SAR data

Abstract In the present study, Sentinel-1A Synthetic Aperture Radar analysis of time series data at C-band was carried out to estimate the winter wheat crop growth parameters. Five different date images were acquired during January 2015–April 2015 at different growth stages from tillering to ripening in Varanasi district, India. The winter wheat crop parameters, i.e. leaf area index, vegetation water content (VWC), fresh biomass (FB), dry biomass (DB) and plant height (PH) were estimated using random forest regression (RFR), support vector regression (SVR), artificial neural network regression (ANNR) and linear regression (LR) algorithms. The Ground Range Detected products of Interferometric Wide (IW) Swath were used at VV polarization. The three different subplots of 1 m2 area were taken for the measurement of crop parameters at every growth stage. In total, 73 samples were taken as the training data-sets and 39 samples were taken as testing data-sets. The highest sensitivity (adj. R2 = 0.95579) of backscattering with VWC was found using RFR algorithm, whereas the lowest sensitivity (adj. R2 = 0.66201) was found for the PH using LR algorithm. Overall results indicate more accurate estimation of winter wheat parameters by the RFR algorithm followed by SVR, ANNR and LR algorithms.

Evaluating the effects of spatial resolution on land use and land cover classification accuracy

The choice of appropriate spatial resolution is a key factor to extract desired information from remotely sensed images. Optical data collected by two different sensors (LISS IV with 5.8 m and Landsat 8-OLI with 30 m spatial resolution respectively) were investigated against the capability to classify accurately into distinct land use and land cover (LULC) classes. To evaluate the quality of training samples class separability analysis using transformed divergence (TD) method was performed. Furthermore, supervised maximum likelihood classifier (MLC) was used to carry out LULC classification. The results indicated that the overall accuracy 83.28% and Kappa coefficient 0.805 for LISS IV image was found higher in comparison to Landsat 8-OLI image having overall accuracy 77.93% and Kappa coefficient 0.742 respectively.

Spatial modeling of SPAD values for different type of crops using LISS-IV satellite imagery

The aim of present study is to assess the feasibility of Linear Imaging Self Scanning (LISS-IV) sensor data for the spatial modeling of SPAD (Soil-Plant Analysis Development) values to monitor the spatial distribution of chlorophyll contents in the agricultural areas. Six crop fields and ten different GPS locations in each crop field were selected for the measurement of SPAD values. The DN (digital number) values of each pixel of LISS-IV satellite image were converted into their top of atmospheric (TOA) reflectance values. The measurement of SPAD values were carried out at the same time of satellite passes over the study area. Two band algorithms was developed using cubic polynomial regression analysis between spectral coefficients and SPAD values. The performance of the two band algorithm was evaluated by using statistical performance indices like %Bias, root mean squared error (RMSE) and Nash-Sutcliffe efficiency (NSE). The values of %Bias, RMSE and NSE were found -0.04, 3.99 and 0.95 respectively.