Dr Sanjay Dudul | Amravati University (original) (raw)
Papers by Dr Sanjay Dudul
DESIGN, CONSTRUCTION, MAINTENANCE, 2021
The Malvaceae family is a kind of plants, which are commonly found near habitat places in India. ... more The Malvaceae family is a kind of plants, which are commonly found near habitat places in India. The earlier approaches reported in the literature were tedious and time consuming with less accuracy due to the similarity in shape and the exine sculptures of pollens. We describe a new classification approach for the classification of three types of pollen grains of the Malvaceae family based on feature vector comprised of Histogram coefficients and image statistics. The approach presented gives precise accuracy in classification of pollen grains of the same family by using SEM images.
To enhance the hiding capacity and security of the well known pixel value differencing (PVD) steg... more To enhance the hiding capacity and security of the well known pixel value differencing (PVD) steganographic method, a new method named Tri-way pixel value differencing (TPVD) was introduced. TPVD is a spatial domain steganographic method in which both horizontal and vertical edges of cover image are utilized for embedding. In TPVD, Use of three different directional edges of cover image for embedding makes it different from PVD, which utilizes only one direction. In this paper a novel method of TPVD steganalysis is proposed, based on the observation that, the difference value of the two sides of a normalized histogram of image under consideration if found to be a positive value other than “0”, then it can be classified as a stego image, where as if this difference is “0” it is a non stego image. The proposed steganalyser can classify test images as stego or cover with 98% accuracy when they contain any secret image. Difference in histogram is plotted for the stego and non stego imag...
2019 International Conference on Recent Advances in Energy-efficient Computing and Communication (ICRAECC)
Computer Networks, Big Data and IoT
Innovations in optical fiber technology are revolutionizing world communications. The focus of th... more Innovations in optical fiber technology are revolutionizing world communications. The focus of this paper is the development of optical fibers that within 20 years displaced copper wire as the transmission medium of choice for most commercial applications in telecommunications systems and computer networks worldwide. High speed and ultra-high capacity of optical communications have emerged as the essential techniques for backbone global information transmission networks. As the bit rate of the transmission system gets higher and higher about 40 Gb/s to 100 Gb/s to several terabits, the modeling of proposed modulation techniques is very important so as to avoid costly practical demonstration. This paper thus describes the various losses associated with fiber and its simulation models and its analysis in OptSim. In this work we will be focusing on chromatic dispersion and fiber induced losses. Initially, it will be observing effects of this chromatic dispersion and fiber induced losse...
Electroencephalography is the most useful and cost effective modality for the diagnosis of epilep... more Electroencephalography is the most useful and cost effective modality for the diagnosis of epilepsy. The detection of these abnormalities by the visual inspection of EEG signals is very complex and time-consuming process and it requires highly skilled doctors. In most of the cases, epilepsy is controlled by the proper medical treatment. For that purpose, the proper and early diagnosis of epilepsy is required. A Clinical Decision Support System (DSS) has been developed for the diagnosis of epilepsy using the Artificial Intelligence Techniques. Different Neural Network based classifiers like MLP, GFFNN, ENN and SVM have been designed and optimized for the diagnosis of epilepsy. In addition, various feature extraction techniques like statistical parameters, Principal Component Analysis, FFT and Discrete Wavelet transform are used for the feature extraction and dimensionality reduction.
WSEAS Transactions on Computers archive, 2018
In tri – way pixel – value differencing(TPVD) steganography approach, three different directional... more In tri – way pixel – value differencing(TPVD) steganography approach, three different directional edges of the cover image are considered to create a stego image. TPVD gives more hiding capacity as compared to original pixel value differencing (PVD) method referring to only one direction. Encryption with steganography gives more secrecy protection to the system. In this paper, a more secure image steganographic technique is proposed, which includes encryption of image before embedding it in, two cover images using TPVD. Experimental results demonstrate that use of two cover images provide high embedding capacity. The peak signal to noise ratio (PSNR), mean square error (MSE) and hiding capacity are evaluated as performance parameter and found to have reasonable values as compared to previous related work. Encryption before embedding protects the information from unauthorized access. The embedded confidential information can be retrieved from the stego images successfully without ass...
Multi –Step ahead prediction of a chaotic time series is a difficult task that has attracted incr... more Multi –Step ahead prediction of a chaotic time series is a difficult task that has attracted increasing interest in recent years. The interest in this work is the development of nonlinear neural network models for the purpose of building multi-step chaotic time series prediction. In the literature there is a wide range of different approaches but their success depends on the predicting performance of the individual methods. Also the most popular neural models are based on the statistical and traditional feed forward neural networks. But it is seen that this kind of neural model may present some disadvantages when long-term prediction is required. In this paper focused time lagged recurrent neural network (FTLRNN) model with gamma memory is developed not only for short-term but also for long-term prediction which allows obtaining better predictions of northern chaotic time series in future. The authors experimented the performance of this FTLRNN model on predicting the dynamic behavi...
The realization of Multiplication operation is essential for most of the scientific applications ... more The realization of Multiplication operation is essential for most of the scientific applications as well as in commercial applications like banking, accounting, financial analysis, tax calculation, currency conversion, and insurance etc. Multiplier is a key building block and almost obligatory component in all such applications. That is why reconfigurable PLDs are equipped with dedicated embedded multipliers. The prerequisite of the multiplier implementation is that it should be primarily fast and secondarily efficient in terms of power consumption and chip area. The multiplication involves two basic operations viz. generation of partial products and their accumulation. To speed up the multiplication process, one can reduce the number of partial products to be generated and later, accelerating their accumulation. The design of multiplier depends upon the type, range and precision of data to be processed by the multiplier block viz. fixed point and floating point numbers represented ...
Epilepsy is one of the major fields of application of EEG. Now a days, identification of epilepsy... more Epilepsy is one of the major fields of application of EEG. Now a days, identification of epilepsy is accomplished manually by skilled neurologist. Those are very small in number. In this work, we propose a methodology for automatic detection of normal, interictal and ictal conditions from recorded of EEG signals. We used the wavelet transform for the feature extraction and obtained statistical parameters from the decomposed wavelet coefficients. The Generalized Feed Forward Neural Network (GFFNN), Multilayer Perceptron (MLP), Elman Neural Network (ENN) and Support Vector Machine (SVM) are used for the classification. The performance of the proposed system was evaluated in terms of classification accuracy, sensitivity, specificity and overall accuracy.
Lecture Notes in Electrical Engineering
Image Steganography involve hiding the secret image in the cover image in such a way that it cann... more Image Steganography involve hiding the secret image in the cover image in such a way that it cannot be detected easily. Hiding the secret image in the selective objects of the cover image is a novel approach that enhances the robustness and security. This paper describes, LSB (Least significant bit) steganography method applied only on certain selected skin tone objects of the cover image to embed the given secret image within. In the proposed work more than one cover image is used to enhance the protection of concealed information. For the skin tone detection of the cover image an alternate colour space YCbCr (Yellow, Chromatic Blue, and Chromatic red) is utilized. Embedding the secret image in the skin region of the cover image provide an excellent protective location for information hiding. Before, that Advance encryption standard technique is used to encrypt secret image which is to be embedded in the selected skin tone objects to have secure stego image. The neural network approach is applied to find the largest object among the selected skin tone objects. The performance parameters like Peak signal to noise ratio and hiding capacity are evaluated thereafter.
2013 International Conference on Advances in Computing, Communications and Informatics (ICACCI), 2013
Palynological data are used in a wide range of applications. A new classification algorithm is pr... more Palynological data are used in a wide range of applications. A new classification algorithm is proposed for pollen grains. With a view to extract features from pollen images, an classifier algorithm is developed which proposes two-dimensional discrete Walsh-Hadamard Transform domain coefficients in addition to image statistics and shape descriptor. The suitability of classifiers based on Multilayer Perceptron (MLP) Neural Network, Generalized Feedforward (GFF) Neural Network, Support Vector Machine (SVM), Radial Basis Functions (RBF) Neural Networks, Recurrent Neural Networks (RNN) and Modular Neural Network (MNN) is explored with the optimization of their respective parameters in view of reduction in time as well as space complexity. Performance of all six classifiers has been compared with respect to MSE, NMSE, and Classification accuracy. The Average Classification Accuracy of MNN comprising of two hidden layers and four parallel MLP neural networks organized in a typical topology is found to be superior (85 % on Cross Validation dataset) amongst all classifiers. Finally, optimal classifier algorithm has been developed on the basis of the best performance. The algorithm suggested could be easily modified to classify more than 10 species. The classifier algorithm will provide an effective alternative to traditional method of pollen image analysis for plant taxonomy and species identification.
IETE Journal of Research, 2007
... 1. INTRODUCTION concerned with the identification of a typical liquid saturated steam heat ex... more ... 1. INTRODUCTION concerned with the identification of a typical liquid saturated steam heat exchanger, which ... It is necessary to establish a ]ayer perceptron) NN (Neural Network) model that can model ... For data and ,t is useful for tuning and simulation before collection, the plant ...
2020 International Conference on Computational Performance Evaluation (ComPE)
Indian Journal of Science and Technology
ABSTRACT In this paper, a novel focused time lagged recurrent neural network (FTLR NN) with gamma... more ABSTRACT In this paper, a novel focused time lagged recurrent neural network (FTLR NN) with gamma memory filter is designed to learn the subtle complex dynamics of a typical magnetic stirrer process. Magnetic stirrer exhibits complex nonlinear operations where reaction is exothermic. It appears to us that identification of such a highly nonlinear system is not yet reported by other researchers using neural networks. As magnetic stirrer process includes time relationship in the input-output mappings, time lagged recurrent neural network is particularly used for identification purpose. The standard back propagation algorithm with momentum term has been proposed in this model. The various parameters like number of processing elements, number of hidden layers, training and testing percentage, learning rule and transfer function in hidden and output layer are investigated on the basis of performance measures like MSE, NMSE and correlation coefficient on testing data set. Finally, effect of different norms are tested along with variation in gamma memory filter. It is shown that dynamic NN model has a remarkable system identification capability for the problem considered in this paper. Thus, FTLR NN with gamma memory filter can be used to learn underlying highly nonlinear dynamics of the system, which is major contribution of this paper. Introduction In any manufacturing process a chemical reactor is at the heart of the plant. Magnetic stirrer is one of the most widely used utility device for pH measurement, reaction rates, distillation and agitation purposes. Magnetic stirrer is a chemical reactor which in size and appearance seems to be one of the least impressive item in manufacturing process but its demand and performance are usually the most important factors in the design of a whole plant. In designing a reactor the aim is to produce a specified product at a given rate from known reactants. As magnetic stirrer exhibit nonlinear operations, performance prediction becomes difficult due to high degree of non-linearity hence exact mathematical modeling is not possible. However, due to development of neural networks, it is possible to develop learning machine based on neural network model that can learn from available experimental data. Thus, a system model can be constructed by estimating unknown plant parameters using neural networks. Inspired from the structure of the human brain and the way it is supposed to be operate, neural networks are parallel computational systems capable of solving number of complex problems in such a diverse areas of as pattern recognition, computer vision, robotics, control and medical diagnosis, to name just few (Haykin, 2003). Neural networks are an effective tool to perform any nonlinear input output mappings. It was the Cybenko (1989) who first proved that, under appropriate conditions, they are able to uniformly approximate any continuous function to any desired degree of accuracy. It is these fundamental results that allow us to employ neural network for system identification purpose. One of the primary reasons for employing neural network was to create a machine that was able to learn from experience. They have the capability to learn the complex nonlinear mappings from a set of observations and predict the future (Dudul, 2007). The present paper carries out neural network based modeling of a typical magnetic stirrer using famous neural network like focused time lag recurrent neural network (FTLR NN) with gamma memory filter. The optimal model is estimated on the basis of performance measures like MSE (Mean Square Error), NMSE (Normalized Mean Square Error), r (Correlation coefficient) and visual inspection of regression characteristics on the testing data sets. Finally it is shown that dynamic NN model has a remarkable system identification capability for the magnetic stirrer process.
Indian Journal of Science and Technology
Journal of Electrical Engineering and Technology, 2009
Ieee Transactions on Instrumentation and Measurement, Oct 1, 2010
Wseas Transactions on Information Science and Applications, 2009
Applied Soft Computing, 2011
A novel method for the classification of material type and its surface roughness by means of a li... more A novel method for the classification of material type and its surface roughness by means of a lightweight plunger probe and optical mouse is presented in this paper. An experimental prototype was developed which involves bouncing or hopping of the plunger based impact probe freely on the plain surface of an object under test. The time and features of bouncing
DESIGN, CONSTRUCTION, MAINTENANCE, 2021
The Malvaceae family is a kind of plants, which are commonly found near habitat places in India. ... more The Malvaceae family is a kind of plants, which are commonly found near habitat places in India. The earlier approaches reported in the literature were tedious and time consuming with less accuracy due to the similarity in shape and the exine sculptures of pollens. We describe a new classification approach for the classification of three types of pollen grains of the Malvaceae family based on feature vector comprised of Histogram coefficients and image statistics. The approach presented gives precise accuracy in classification of pollen grains of the same family by using SEM images.
To enhance the hiding capacity and security of the well known pixel value differencing (PVD) steg... more To enhance the hiding capacity and security of the well known pixel value differencing (PVD) steganographic method, a new method named Tri-way pixel value differencing (TPVD) was introduced. TPVD is a spatial domain steganographic method in which both horizontal and vertical edges of cover image are utilized for embedding. In TPVD, Use of three different directional edges of cover image for embedding makes it different from PVD, which utilizes only one direction. In this paper a novel method of TPVD steganalysis is proposed, based on the observation that, the difference value of the two sides of a normalized histogram of image under consideration if found to be a positive value other than “0”, then it can be classified as a stego image, where as if this difference is “0” it is a non stego image. The proposed steganalyser can classify test images as stego or cover with 98% accuracy when they contain any secret image. Difference in histogram is plotted for the stego and non stego imag...
2019 International Conference on Recent Advances in Energy-efficient Computing and Communication (ICRAECC)
Computer Networks, Big Data and IoT
Innovations in optical fiber technology are revolutionizing world communications. The focus of th... more Innovations in optical fiber technology are revolutionizing world communications. The focus of this paper is the development of optical fibers that within 20 years displaced copper wire as the transmission medium of choice for most commercial applications in telecommunications systems and computer networks worldwide. High speed and ultra-high capacity of optical communications have emerged as the essential techniques for backbone global information transmission networks. As the bit rate of the transmission system gets higher and higher about 40 Gb/s to 100 Gb/s to several terabits, the modeling of proposed modulation techniques is very important so as to avoid costly practical demonstration. This paper thus describes the various losses associated with fiber and its simulation models and its analysis in OptSim. In this work we will be focusing on chromatic dispersion and fiber induced losses. Initially, it will be observing effects of this chromatic dispersion and fiber induced losse...
Electroencephalography is the most useful and cost effective modality for the diagnosis of epilep... more Electroencephalography is the most useful and cost effective modality for the diagnosis of epilepsy. The detection of these abnormalities by the visual inspection of EEG signals is very complex and time-consuming process and it requires highly skilled doctors. In most of the cases, epilepsy is controlled by the proper medical treatment. For that purpose, the proper and early diagnosis of epilepsy is required. A Clinical Decision Support System (DSS) has been developed for the diagnosis of epilepsy using the Artificial Intelligence Techniques. Different Neural Network based classifiers like MLP, GFFNN, ENN and SVM have been designed and optimized for the diagnosis of epilepsy. In addition, various feature extraction techniques like statistical parameters, Principal Component Analysis, FFT and Discrete Wavelet transform are used for the feature extraction and dimensionality reduction.
WSEAS Transactions on Computers archive, 2018
In tri – way pixel – value differencing(TPVD) steganography approach, three different directional... more In tri – way pixel – value differencing(TPVD) steganography approach, three different directional edges of the cover image are considered to create a stego image. TPVD gives more hiding capacity as compared to original pixel value differencing (PVD) method referring to only one direction. Encryption with steganography gives more secrecy protection to the system. In this paper, a more secure image steganographic technique is proposed, which includes encryption of image before embedding it in, two cover images using TPVD. Experimental results demonstrate that use of two cover images provide high embedding capacity. The peak signal to noise ratio (PSNR), mean square error (MSE) and hiding capacity are evaluated as performance parameter and found to have reasonable values as compared to previous related work. Encryption before embedding protects the information from unauthorized access. The embedded confidential information can be retrieved from the stego images successfully without ass...
Multi –Step ahead prediction of a chaotic time series is a difficult task that has attracted incr... more Multi –Step ahead prediction of a chaotic time series is a difficult task that has attracted increasing interest in recent years. The interest in this work is the development of nonlinear neural network models for the purpose of building multi-step chaotic time series prediction. In the literature there is a wide range of different approaches but their success depends on the predicting performance of the individual methods. Also the most popular neural models are based on the statistical and traditional feed forward neural networks. But it is seen that this kind of neural model may present some disadvantages when long-term prediction is required. In this paper focused time lagged recurrent neural network (FTLRNN) model with gamma memory is developed not only for short-term but also for long-term prediction which allows obtaining better predictions of northern chaotic time series in future. The authors experimented the performance of this FTLRNN model on predicting the dynamic behavi...
The realization of Multiplication operation is essential for most of the scientific applications ... more The realization of Multiplication operation is essential for most of the scientific applications as well as in commercial applications like banking, accounting, financial analysis, tax calculation, currency conversion, and insurance etc. Multiplier is a key building block and almost obligatory component in all such applications. That is why reconfigurable PLDs are equipped with dedicated embedded multipliers. The prerequisite of the multiplier implementation is that it should be primarily fast and secondarily efficient in terms of power consumption and chip area. The multiplication involves two basic operations viz. generation of partial products and their accumulation. To speed up the multiplication process, one can reduce the number of partial products to be generated and later, accelerating their accumulation. The design of multiplier depends upon the type, range and precision of data to be processed by the multiplier block viz. fixed point and floating point numbers represented ...
Epilepsy is one of the major fields of application of EEG. Now a days, identification of epilepsy... more Epilepsy is one of the major fields of application of EEG. Now a days, identification of epilepsy is accomplished manually by skilled neurologist. Those are very small in number. In this work, we propose a methodology for automatic detection of normal, interictal and ictal conditions from recorded of EEG signals. We used the wavelet transform for the feature extraction and obtained statistical parameters from the decomposed wavelet coefficients. The Generalized Feed Forward Neural Network (GFFNN), Multilayer Perceptron (MLP), Elman Neural Network (ENN) and Support Vector Machine (SVM) are used for the classification. The performance of the proposed system was evaluated in terms of classification accuracy, sensitivity, specificity and overall accuracy.
Lecture Notes in Electrical Engineering
Image Steganography involve hiding the secret image in the cover image in such a way that it cann... more Image Steganography involve hiding the secret image in the cover image in such a way that it cannot be detected easily. Hiding the secret image in the selective objects of the cover image is a novel approach that enhances the robustness and security. This paper describes, LSB (Least significant bit) steganography method applied only on certain selected skin tone objects of the cover image to embed the given secret image within. In the proposed work more than one cover image is used to enhance the protection of concealed information. For the skin tone detection of the cover image an alternate colour space YCbCr (Yellow, Chromatic Blue, and Chromatic red) is utilized. Embedding the secret image in the skin region of the cover image provide an excellent protective location for information hiding. Before, that Advance encryption standard technique is used to encrypt secret image which is to be embedded in the selected skin tone objects to have secure stego image. The neural network approach is applied to find the largest object among the selected skin tone objects. The performance parameters like Peak signal to noise ratio and hiding capacity are evaluated thereafter.
2013 International Conference on Advances in Computing, Communications and Informatics (ICACCI), 2013
Palynological data are used in a wide range of applications. A new classification algorithm is pr... more Palynological data are used in a wide range of applications. A new classification algorithm is proposed for pollen grains. With a view to extract features from pollen images, an classifier algorithm is developed which proposes two-dimensional discrete Walsh-Hadamard Transform domain coefficients in addition to image statistics and shape descriptor. The suitability of classifiers based on Multilayer Perceptron (MLP) Neural Network, Generalized Feedforward (GFF) Neural Network, Support Vector Machine (SVM), Radial Basis Functions (RBF) Neural Networks, Recurrent Neural Networks (RNN) and Modular Neural Network (MNN) is explored with the optimization of their respective parameters in view of reduction in time as well as space complexity. Performance of all six classifiers has been compared with respect to MSE, NMSE, and Classification accuracy. The Average Classification Accuracy of MNN comprising of two hidden layers and four parallel MLP neural networks organized in a typical topology is found to be superior (85 % on Cross Validation dataset) amongst all classifiers. Finally, optimal classifier algorithm has been developed on the basis of the best performance. The algorithm suggested could be easily modified to classify more than 10 species. The classifier algorithm will provide an effective alternative to traditional method of pollen image analysis for plant taxonomy and species identification.
IETE Journal of Research, 2007
... 1. INTRODUCTION concerned with the identification of a typical liquid saturated steam heat ex... more ... 1. INTRODUCTION concerned with the identification of a typical liquid saturated steam heat exchanger, which ... It is necessary to establish a ]ayer perceptron) NN (Neural Network) model that can model ... For data and ,t is useful for tuning and simulation before collection, the plant ...
2020 International Conference on Computational Performance Evaluation (ComPE)
Indian Journal of Science and Technology
ABSTRACT In this paper, a novel focused time lagged recurrent neural network (FTLR NN) with gamma... more ABSTRACT In this paper, a novel focused time lagged recurrent neural network (FTLR NN) with gamma memory filter is designed to learn the subtle complex dynamics of a typical magnetic stirrer process. Magnetic stirrer exhibits complex nonlinear operations where reaction is exothermic. It appears to us that identification of such a highly nonlinear system is not yet reported by other researchers using neural networks. As magnetic stirrer process includes time relationship in the input-output mappings, time lagged recurrent neural network is particularly used for identification purpose. The standard back propagation algorithm with momentum term has been proposed in this model. The various parameters like number of processing elements, number of hidden layers, training and testing percentage, learning rule and transfer function in hidden and output layer are investigated on the basis of performance measures like MSE, NMSE and correlation coefficient on testing data set. Finally, effect of different norms are tested along with variation in gamma memory filter. It is shown that dynamic NN model has a remarkable system identification capability for the problem considered in this paper. Thus, FTLR NN with gamma memory filter can be used to learn underlying highly nonlinear dynamics of the system, which is major contribution of this paper. Introduction In any manufacturing process a chemical reactor is at the heart of the plant. Magnetic stirrer is one of the most widely used utility device for pH measurement, reaction rates, distillation and agitation purposes. Magnetic stirrer is a chemical reactor which in size and appearance seems to be one of the least impressive item in manufacturing process but its demand and performance are usually the most important factors in the design of a whole plant. In designing a reactor the aim is to produce a specified product at a given rate from known reactants. As magnetic stirrer exhibit nonlinear operations, performance prediction becomes difficult due to high degree of non-linearity hence exact mathematical modeling is not possible. However, due to development of neural networks, it is possible to develop learning machine based on neural network model that can learn from available experimental data. Thus, a system model can be constructed by estimating unknown plant parameters using neural networks. Inspired from the structure of the human brain and the way it is supposed to be operate, neural networks are parallel computational systems capable of solving number of complex problems in such a diverse areas of as pattern recognition, computer vision, robotics, control and medical diagnosis, to name just few (Haykin, 2003). Neural networks are an effective tool to perform any nonlinear input output mappings. It was the Cybenko (1989) who first proved that, under appropriate conditions, they are able to uniformly approximate any continuous function to any desired degree of accuracy. It is these fundamental results that allow us to employ neural network for system identification purpose. One of the primary reasons for employing neural network was to create a machine that was able to learn from experience. They have the capability to learn the complex nonlinear mappings from a set of observations and predict the future (Dudul, 2007). The present paper carries out neural network based modeling of a typical magnetic stirrer using famous neural network like focused time lag recurrent neural network (FTLR NN) with gamma memory filter. The optimal model is estimated on the basis of performance measures like MSE (Mean Square Error), NMSE (Normalized Mean Square Error), r (Correlation coefficient) and visual inspection of regression characteristics on the testing data sets. Finally it is shown that dynamic NN model has a remarkable system identification capability for the magnetic stirrer process.
Indian Journal of Science and Technology
Journal of Electrical Engineering and Technology, 2009
Ieee Transactions on Instrumentation and Measurement, Oct 1, 2010
Wseas Transactions on Information Science and Applications, 2009
Applied Soft Computing, 2011
A novel method for the classification of material type and its surface roughness by means of a li... more A novel method for the classification of material type and its surface roughness by means of a lightweight plunger probe and optical mouse is presented in this paper. An experimental prototype was developed which involves bouncing or hopping of the plunger based impact probe freely on the plain surface of an object under test. The time and features of bouncing