vs jayanthi

SAVEETHA UNIVERSITY, Electronics and Communication Engineering, Faculty Member

Followers

Following

Public Views

Interests

Uploads

Papers by vs jayanthi

Non-Small Cell Lung Tumor Identification by CNN and ANN Classifier

Psychology and Education Journal, Feb 1, 2021

Alzheimer’s Disease Classification Using Deep Convolutional Neural Network

2019 9th International Conference on Advances in Computing and Communication (ICACC), 2019

Alzheimer’s Disease is the most common form of dementia which initially destroys the memory and f... more Alzheimer’s Disease is the most common form of dementia which initially destroys the memory and finally progresses to death. This irreversible disease is mostly found among older people. The latest innovations on the multimodal neuroimaging data made it possible to detect the disease in life which was a major breakthrough in neuroscience. However, the larger degree of similarity between the brain images was the major challenge in the diagnosis. The Deep Learning technique has gained excellent results on image classification among the present researches. Hence it is utilized for the classification of brain images among Cognitively Normal (CN), Early Mild Cognitive Impairment (EMCL), Mild Cognitive Impairment (MCL), Late Mild Cognitive Impairment (LMCI), Alzheimer’s Disease (AD) which are the five classes of AD thus ensuring very precise and accurate diagnosis. The transfer learning approach has been taken up for the classification process by which three pre-trained networks, namely AlexNet, ResNet-18 and, GoogLe Net are modified and trained for 3000 images. All the three networks are trained for the same set of images which were acquired from the ADNI database.

Research Article Low Power Adder Based Auditory Filter Architecture

Attribution License, which permits unrestricted use, distribution, and reproduction in any medium... more Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Cochlea devices are powered up with the help of batteries and they should possess long working life to avoid replacing of devices at regular interval of years. Hence the devices with low power consumptions are required. In cochlea devices there are numerous filters, each responsible for frequency variant signals, which helps in identifying speech signals of different audible range. In this paper, multiplierless lookup table (LUT) based auditory filter is implemented. Power aware adder architectures are utilized to add the output samples of the LUT, available at every clock cycle. The design is developed and modeled using Verilog HDL, simulated using Mentor Graphics Model-Sim Simulator, and synthesized using Synopsys Design Compiler tool. The design was mapped to TSMC 65 nm technological node. The standard ASIC design methodology has been ad...

A Local Topology Reconfiguration Mechanism to Improve the Energy Efficiency in Wireless Sensor Network

A wireless sensor network (WSN) is a self organizing wireless network system consists of number o... more A wireless sensor network (WSN) is a self organizing wireless network system consists of number of energy limited sensor nodes. Since sensor nodes are battery powered devices, they have limited processing and transmission power and with limited communication ability. In order to transmit sensing data to receiver effectively, it is necessary to design an energy efficient routing protocol for WSNs. The nodes with greater signal strength will have more communication link and result in faster energy consumption as a whole network cannot always work under the same topology structure. A topology reconnecting mechanism of the cluster head rotation algorithm is needed. This paper presents an Enhanced Forward Aware Factor -Energy Balanced Routing Method (EFAF-EBRM) based on the selection of appropriate cluster head to improve the energy efficiency . In this technique the sensor node may prefer to select cluster head that makes the total energy consumption least.This proposed cluster formatio...

Download

Pleural Plaque Detection Using Hybrid Segmentation Algorithm and Elm Classification

Pleural plaques are by far the most common indication of significant exposure to asbestos and act... more Pleural plaques are by far the most common indication of significant exposure to asbestos and act as a biomarker for the diagnosis of lung cancer in later stages. Plaques can develop on both layers of the pleura, a thin membrane that surrounds the lungs and aids in breathing. They most commonly develop on the parietal pleura, which lines the inside of the rib cage, but can also affect the visceral pleura, which lines the lungs. For lung image segmentation, many clustering and threshold techniques have been proposed. Here initially the image is pre-processed using anisotropic diffusion filter. Then pleural plaque is detected and segmented using region growing approach followed by layer refinement using active contour model (ACM). Finally the segmented image is classified so as to yield normal and abnormal set using Extreme Learning Machine (ELM). Segmentation is carried using chest CT images to determine the efficiency of proposed technique.

Download

Enhanced Security of Sensitive Images by Cryptography and Hiding Methods

2018 3rd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology (RTEICT), 2018

Natural image based Visual Cryptography is an efficient technique used to secure sensitive images... more Natural image based Visual Cryptography is an efficient technique used to secure sensitive images. Here images are hidden inside shares which appear as noise-like images. Now a days utmost concern is given to the privacy of images which includes fingerprint, signature and other secret information. This scheme suffers a high transmission risk since the secret images are hidden inside shares. In this paper, the generated share is firstly hidden in a printed media using QR technology and secondly in a cover media with the help of steganographic techniques. It enhances the security and reduces the risk of eavesdropping in the transmission phase. Experimental results show that steganography performs better. The average PSNR values of Steganography and QR technology are 26.81 dB and 22.45 dB respectively.

A Literature Survey on 3-D Optoacoustic Tomography and Its Reconstruction Patterns

Optoacoustic tomography (OAT) is a hybrid, non invasive method which is also known as photo acous... more Optoacoustic tomography (OAT) is a hybrid, non invasive method which is also known as photo acoustic tomography, is an emerging computed biomedical imaging modality. Reconstruction method is used to reconstruct the OAT images, among which iterative reconstruction algorithm is the one that is designed to minimize the error between measured signals and signals from the reconstructed image. The Kaiser-Bessel window functions for iterative reconstruction are accomplished for the accurate computation of the system matrix. In this work we review the different types of reconstruction algorithm such as pseudo inverse reconstruction, filtered back projection, simple back projection and direct Fourier Transform method.

Download

Noise Reduction Algorithms for Better Speech Understanding in Cochlear Implantation- a Survey

Cochlear Implant (CI) is an electronic device which provides a sense of sound to a person who is ... more Cochlear Implant (CI) is an electronic device which provides a sense of sound to a person who is deaf by birth or severely hard of hearing. Improvement in speech understanding in noisy environment is a major challenge in cochlear implant. Various noise reduction algorithms have been designed in cochlear implant for better understanding in noisy environment. Starting from Two microphone adaptive beamformer method, SDW-MWF, SPMWF, EVD based filter and GEVD based approximation of the filters are discussed here. Survey regarding the various algorithms for noise reduction and better speech understanding performance in cochlear implants are studied and analysed. KeywordsNR algorithms, SDW-MWF, EVD, SP-MWF, and GEVD based NR filters

Download

Design of Vector Quantizer for Image Compression Using Hybrid ANT GLA algorithm

Compression is a key technique for digital transmission and storage of images due to the vast amo... more Compression is a key technique for digital transmission and storage of images due to the vast amount of data associated with it. Image compression algorithms using Vector Quantization (VQ) have been receiving considerable attention. In VQ, minimization of Mean Square Error (MSE) between code book vectors and training vectors is a non-linear problem. Traditional GLA algorithm used for designing the codebooks for Vector Quantizer converges to a local minimum, which depends on the initial code book. Ant colony algorithm is a population-based meta-heuristic search approach that has been receiving increasing attention in the recent years. Ant colony algorithm is inspired by the behavior of real ant colonies. It has shown to be successful and popular for solving optimization problems. In this paper we present a new hybrid approach in which Ant colony algorithm is combined with GLA to design a codebook for vector quantization. Simulation results indicate that ANTGLA has better performance ...

Review of Image Encryption Technique Based on Aes

Encryption of images has become a necessity in today’s world to protect a confidential image from... more Encryption of images has become a necessity in today’s world to protect a confidential image from unauthorized access. Various methods have been proposed and developed in spatial, frequency and hybrid domains to encrypt the image securely. Encryption could either be a full encryption or partial encryption based on the security requirements. In this paper we present an overview of various image encryption techniques based on AES. KeywordsAES, Image Encryption, Security, Cryptography

Download

Lung Nodule Classification in CT Images Using Convolutional Neural Network

2019 9th International Conference on Advances in Computing and Communication (ICACC), 2019

Convolutional neural networks (CNN), due to its self-learning ability and deep feature extraction... more Convolutional neural networks (CNN), due to its self-learning ability and deep feature extraction capability has created a new era to the biomedical applications and medical image analysis. Computer aided diagnosis system (CAD) using CNN’s helps radiologists in decision making and reduces their difficulty on screening lung nodules. This paper presents CAD system using pre-trained networks AlexNet, VGG16 and a new network is designed based on the inference obtained to eliminate the drawbacks of available network for the early prediction of lung cancer. Comparative analysis based on the performance parameters proved that the proposed CNN architecture outperforms the pre-trained networks. Experiments were evaluated using computed tomography (CT) scan images obtained from LIDC-IDRI dataset.

Design of Low Power Speech Processor-Based Cochlear Implants Using Modified FIR Filter with Variable Frequency Mapping

Journal of Circuits, Systems and Computers, 2020

Generally, cochlear implant system consists of one or more electrodes which directly activate the... more Generally, cochlear implant system consists of one or more electrodes which directly activate the auditory nerve. The microphone transforms external speech processor input into stimuli for each electrode and transcutaneous the connection between the electrodes and processor. As a result, cochlear implants continue to improve the performance of speech processor. After the implantation, preservation gains importance in acoustic hearing to prevent the neural degradation from the loss of cells. Eventually, the possibility of preserving low-frequency and high-frequency hearing using modified surgical technique has been explored for hearing in cochlear implantation. The primary focus on the architecture is to reduce the major potential thread performance and power trade-off. So, frequency mapping function with modified FIR filter has been proposed. For this implementation, the values relative to the frequency range defines upper and lower stimulus limits of each electrode. The proposed FI...

Multiple share creation based visual cryptographic scheme using diffusion method with a combination of chaotic maps for multimedia applications

Multimedia Tools and Applications, 2019

In the modern years, multimedia security has become a prime agenda for secure transmission over u... more In the modern years, multimedia security has become a prime agenda for secure transmission over unsecured channels due to the increase in digitalization of media. Cryptography is one of the techniques, which can be used for security of exchanged data. The visual cryptographic scheme generates numerous unreasonable shares of the image with specific information. The secret information hidden in the image can be retrieved by accumulating all the shares. The basic idea behind this cryptographic scheme is the encryption of image which hides the secret into m several image shares. By this process, it turns to be complex for the hackers to retrieve the original data of the image. Furthermore, to increase the confidentiality and privacy of the original image, there is in demand to make use of the efficient encryption algorithm. According to this, a new image encryption and decryption is developed using diffusion method with a combination of chaotic maps. Overall, the proposed visual cryptographic method is divided into three phases namely, i) Separation of color bands, ii) Generation of several shares and iii) Encryption & Decryption. Firstly, the color image is divided into three bands (R, G and B) and then various shares of the image are generated based on the pixel measures. After that, the encryption and decryption are done by using diffusion process associated with chaotic map. Experimental results of various analysis and computer simulation confirm that the new algorithm offers high security and is suitable for practical image encryption.

Download

A Research on Cluster Based Routing Methodologies for Wireless Sensor Networks

International Journal of Trend in Scientific Research and Development, 2017

The design of Wireless sensor networks has currently significant attention due to their limitless... more The design of Wireless sensor networks has currently significant attention due to their limitless potential and has great value of applications such as not closely connected environmental monitoring and target tracking. This has been motivated by the development and availability of small, low cost and intelligent sensors. A wireless sensor network is a collection of sensor nodes prearranged into a cooperative network. These sensors nodes are outfitted with limited processing and computing resources. Since radio transmission and reception consumes a lot of power, one of the critical suspicions in wireless sensor network is the inherent limited battery power within the sensor nodes. Therefore battery power is one of the vital parameter for an algorithm design to increase the lifespan of the sensor node. Recent researches have been carried out in investigating diverse aspects in connection with low power routing protocols. In this paper, the focus is mainly driven towards the review of energy efficient cluster-based available routing protocols for wireless sensor network. Also a methodology by which optimal routing can be achieved in the context of energy is portrayed artistically.

Download

An efficient classification approach for detection of Alzheimer’s disease from biomedical imaging modalities

Multimedia Tools and Applications, 2018

The complex patterns of the neuroimaging data are analyzed successfully with biomedical imaging a... more The complex patterns of the neuroimaging data are analyzed successfully with biomedical imaging applications. The patients with/without AD can be discriminated effectively through several biomedical imaging modalities such as, sMRI, fMRI, PET and so on. In this paper, brain images from structural MRI (sMRI) are used for better categorization of 3 subjects namely, NC (Normal Control), MCI (Mild Cognitive Impairment) and AD (Alzheimer's disease). Moreover, ambiguous training data employed for the discrimination of subjects may mislead the classifier to take incorrect decisions and in turn degrades the classification performance. In order to recover these hurdles, we propose an automated reliable system for the detection of AD affected patients accurately from the brain images of sMRI. The proposed system is a multi-stage system comprising four key phases namely, i) pre-processing, ii) feature extraction, iii) feature selection and iv) detection phase. In the initial phase, ROI regions related to Hippocampus (HC) and Posterior Cingulate Cortex (PCC) from the brain images are extracted using Automated Anatomical Labeling (AAL) method. In the feature extraction stage, important texture and shape features are extracted from HC and PCC involved in three brain planes. Nearly, 19 highly relevant AD related features are selected through a multiple-criterion feature selection method. It should be noted that, the class labels when explored manually consumes more time and turns to be an expensive process. Therefore, it is essential to construct an automatic method to identify irrelevant samples in the training data to enhance the decision-making process. With this in mind, at the detection phase, a novel classification technique is proposed by combining the Kernel fuzzy c-means clustering (i.e. unsupervised learning technique) and Back-propagation artificial neural network (i.e. supervised learning technique) to categorize NC, MCI and AD from the brain images of sMRI. This proposed KFCM based BPANN algorithm can improve the classification performance by removing the suspicious training samples. The proposed frameworks efficiency is evaluated Multimed Tools Appl

Download

Optimal visual cryptographic scheme with multiple share creation for multimedia applications

Computers & Security, 2018

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service... more This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. Highlights  An new approach to improve privacy and confidentiality for multimedia data-security  OGWO-based iterative process for the generation of optimal keys  Opposition base learning (OBL) process is used to enhance the conventional GWO  Detailed experimental result is analyzed with the existing approaches.

Download

Extraction of myocardial wall from cardiac CT images

2015 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS), 2015

Congenital heart diseases like Atrial Septal Defect, Ventricular Septal Defect and plaque can be ... more Congenital heart diseases like Atrial Septal Defect, Ventricular Septal Defect and plaque can be found by delineation of myocardial wall. Cardiac Computed Tomography is a non invasive imaging technique used for diagnosis of coronary artery diseases. Cardiac chamber segmentation is particularly valuable for the extraction of myocardial wall, because changes in the thickness of myocardium are used to detect the abnormality in heart functions. In this paper initially image is pre-processed using median filter and edge detection operator. The Left Ventricle and Right Ventricle are segmented using two algorithms, Region Growing and Active Contour Model. Moving window concept is introduced for seed selection and mask initialization. Relevant Gray Level Co-occurrence Matrix features are extracted for classification purpose. Finally the segmented image is classified using Adaptive Neural Fuzzy Inference System classifier.

Iterative Kernel PCA based Classification of Retinal Images for Diabetic Macular Edema

International Journal of Computer Applications, 2013

Swelling in the macular region of retina which is also known as macular edema, is a complication ... more Swelling in the macular region of retina which is also known as macular edema, is a complication of the eye often leading to reduced capacity of vision. Diabetic retinopathy is also a severe complication to vision. In this work, iterative kernel based PCA is proposed which is a novel method used for the classification purpose in diseased retinal images. Exudate detection is carried out via a supervised learning approach using the normal fundus images. Feature extraction is introduced to capture the global characteristics of the fundus images and discriminate the normal from diseased images. The performance of the proposed methodology with the conventional PCA is evaluated based on classification accuracy. Experimental results shows the superior nature of iterative kernel based PCA in terms of performance measures.

Download

A new architecture for the generation of picture based CAPTCHA

2011 3rd International Conference on Electronics Computer Technology, 2011

ABSTRACT Automated network attack such as denial-of-service (DoS) leads to significant wastage of... more ABSTRACT Automated network attack such as denial-of-service (DoS) leads to significant wastage of resources, which is a common threat to network security. To prevent these automated network attacks CAPTCHA based security mechanism is to be adopted so that it will differentiate humans from machines. Optical Character Recognition (OCR) based CAPTCHAs are more vulnerable to automated attacks due to the existence of correlation algorithms and direct distortion estimation techniques. The illegibility of the text CAPTCHA makes the user difficult to read it and thus they feel uncomfortable .In order to overcome these difficulties a new type of CAPTCHA ,that is, picture based CAPTCHAs came into existence ,which are more efficient and secure than the existing text based CAPTCHAs. We propose a new architecture for the generation of picture based CAPTCHA, which is resistant to segmentation through edge detection and thresholding,shape matching and random guessing. Our security analysis shows that the proposed architecture is showing better results in comparison with other picture based CAPTCHAs.

Low Power Adder Based Auditory Filter Architecture

The Scientific World Journal, 2014

Cochlea devices are powered up with the help of batteries and they should possess long working li... more Cochlea devices are powered up with the help of batteries and they should possess long working life to avoid replacing of devices at regular interval of years. Hence the devices with low power consumptions are required. In cochlea devices there are numerous filters, each responsible for frequency variant signals, which helps in identifying speech signals of different audible range. In this paper, multiplierless lookup table (LUT) based auditory filter is implemented. Power aware adder architectures are utilized to add the output samples of the LUT, available at every clock cycle. The design is developed and modeled using Verilog HDL, simulated using Mentor Graphics Model-Sim Simulator, and synthesized using Synopsys Design Compiler tool. The design was mapped to TSMC 65 nm technological node. The standard ASIC design methodology has been adapted to carry out the power analysis. The proposed FIR filter architecture has reduced the leakage power by 15% and increased its performance by 2.76%.

Download

vs jayanthi

Uploads

Papers by vs jayanthi

Log In