Volume 6 Issue 4
Computer-Aided Design for skin disease Identification and Categorization Using Image Processing
Author's Name: G.Charulatha, R.Dineskkumar, B.Kaleeswari & K.Lakshmipriya

Abstract—One of the many significant prevalent illnesses is skin disease. Features extraction is essential for supporting the categorization of skin illnesses in skin disease detection. The skill of the physicians and the results of skin function tests are particularly dependent on the capacity to diagnose skin illness, which is a laborious process. To improve diagnosis accuracy and prevent inadequate human sources, an automated computer-aided design (CAD) for skin disease identification and categorization via photographs is required. A critical duty is identifying skin conditions from a picture. It mostly depends on the accurately measured and arranged characteristics of the illnesses. Creating usable photos is made more difficult by the same visual characteristics shared by several skin conditions. The diagnosis would be improved by formal investigation into these disorders from the picture.

B&W Image Colorization Using Computer Vision And Deep Learning
Author's Name: Dr Palson Kennedy R, Jonisha S, Bersikin Libina T, Vasantha Raja S S

Abstract—Previous approaches to black and white image colorization relied on manual human annotation, which frequently resulted in desaturated results that were not believable as true colorizations. The project attempts to generate a plausible color version of a greyscale photograph given as input. It's a fully automated process for creating beautiful, lifelike colorization. By framing the challenge as a classification job, it accepts the problem's underlying ambiguity and uses class re-balancing during training to increase the diversity of colors in the end result. The system is trained on over a million color images and is implemented as a feed-forward pass in a CNN during testing. By a large margin, this strategy surpasses earlier methods. It also shows that colorization can be an effective pretext task for self-supervised feature learning when employed as a cross-channel encoder. On a variety of feature learning benchmarks, this technique achieves cutting-edge results

An Investigation of Low Power VLSI Design Techniques
Author's Name: R.Dinesh Kumar, M.Ramkumar Prabhu, K.Lakshmi Priya, M.Renuga, K.S.Senthil Kumar, V.Vennisa

Abstract— low power has become a major subject in the electronics industries of today. For the design of VLSI chips, power dissipation has taken on equal importance to performance and area. The main issues below 90nm due to increased complexity are lowering power usage and overall power management on chip. Due to the requirement to lower package costs and increase battery life, power optimization is crucial for many systems. In low power VLSI designs, leakage current also has a significant impact on power management. An growing portion of integrated circuits' overall power dissipation is being accounted for by leakage current. This paper discusses numerous power management techniques, methodologies, and tactics for low power circuits and systems. Future challenges for designing low power high performance circuits are also discussed

An Integrated Motor-Drive and Battery-Charging System Based on a Split-Field-Winding Doubly Salient Electromagnetic Machine
Author's Name: A.Antonycharles, Dr.S.Dinakar Raj, S.L.Sreedevi R.Tamilamuthan, Dr.G.Irin Loretta, A.Vijayalakshmi

Abstract— An integrated motor-drive and battery charging (IMB) system based on the BLDC is proposed in this paper, where field winding of is integrated as the filter inductance of the front-end DC/DC converter. In the driving mode, the coupling relationship between the front-end DC/DC converter and rear-end inverter for BLDC is deduced, and the corresponding basic control method is given to achieve the independent control of the front-end DC/DC converter and rear-end inverter based on the power balance of the cascade converter. While, due to the constraints analysis of the IMB system in the dynamic procedure, an optimized tri-loop control strategy of speed, power and dual-threshold voltage is proposed to improve output performance. Finally, the simulation model and experiment platform are established to verify the effectiveness and the steady and dynamic performance of the IMB system with the proposed driving control strategy.

Analysis of Formability of Aa6061 by Dual Step Angle Single Point Incremental Forming
Author's Name: Sivakumar.R, Prabhakaran, Sathiyamoorthy.R.M

Abstract— Single Point Incremental Forming (SPIF) is one among the chief promising more current strategies in metal shaping procedure, and this procedure depends on confined plastic misshapening on a thin level strong clear. Steady framing process has higher formability limits than other level strong shaping procedures including stepping, and is in this way an attractive technique for shaping level strong parts. to require bit of leeway of this high formability it's important to realize an approach to boost the formability, improving the probability of segment achievement. It distorts continuously and locally the level strong by round framing. Hemi-round headed device is to be acclimated distort the sheet in to required shapes. Formability investigation must be dispensed to get effective shaping procedure. This venture plans to survey the framing conduct of AA6061 through steady shaping procedure by double step framing edge. Shortened cones are to be framed for this reason at temperature, with the assistance of PC controlled numerical machine Formability Limit Diagram (FLD) is utilized to survey the conduct of level strong. FLD and thickness circulation has been anticipated and looked at for the structure.

Design of an MLI with a reduced amount of Switches
Author's Name: A.Vijayalakshmi, Dr.P.Yamunaa, Mr.A.Antonycharles, Dr.S.Dinakar Raj, S.L. Sreedevi, R.Tamilamuthan

Abstract— In this paper, 31-level symmetric inverter is designed using optimal number of switches which produces higher output voltage levels with low harmonic distortion. The Proposed multilevel inverter output voltage level increasing by using eight numbers of switches driven by the multicarrier modulation techniques. The voltage sources used in this multilevel CASCADE inverter is asymmetric in nature to generate output voltage with reduced distortion. By using six asymmetric voltage sources and eight switches, 31-level inverter is simulated and the results of the inverter topology are studied in view of reduced harmonic components. This proposed topology offers high power capability associated with less commutation losses, less total harmonic distortion (THD), and involves less number of switching devices and less number of voltage source in comparison of conventional topologies for 31-Level asymmetric multilevel inverter.

Design of Smart Helmet
Author's Name: Sivakumar.R, Prabhakaran.D, Sathiyamoorthy.R.M.

Abstract— In present time many cases of bike accident can be seen around us. peoples get injured or might be dead and one of the reason is not wearing helmet. Many peoples could save their life in accident cases if they weared helmet at the time of accident, continuously road rules are violated. So as to overcome these problems, an Smart helmet is proposed having a control system built on a helmet. Smart Helmet for Motorcyclist is a project undertaken to increase the rate of road safety among motorcyclists. The idea is obtained after knowing that the increasing number of fatal road accidents over the years is cause for concern among motorcyclists. A smart helmet is a type of protective headgear used by the rider which makes bike driving safer than before. The main purpose of this helmet is to provide safety for the rider. It will send the message and dial call which can be easy to connect the rider with the emergency number when he/she met an accident with helmet on. A smart helmet is a special idea which makes motorcycle driving safer than before. This is implemented using GSM module technology. The working of this smart helmet is very simple, vibration sensor sense the vibration of helmet where the probability of hitting is more which are connected to nano-arduino board. So when the rider crashes and the helmet hit the ground, these sensors sense and gives to the nano-arduino board, When the data exceeds minimum stress limit then GSM module automatically sends message and call to Emergency or family members. Also here we have an fan for rider’s comfort and light indications to indicate other vehicles. May Wearing this smart helmet can save a life.

Google Playstore Reviews Prediction Using ML And NLP
Author's Name: Vijayanarayanan A, Savithiri R, Lekha P, Abbirami R S

Abstract— Play Store is Google's official pre-installed app store on Android-certified devices. It provides access to content on the Google Play Store, including apps, books, magazines, and music, movies, and television programs. Google play store allow the user to download a mobile application and user get inspired by the rating and reviews of the mobile app. A recent study analyzes that user preferences, user opinion for improvement, user sentiment about particular feature and detail with descriptions of experiences are very useful for an application development. The aim is to classify the Google app reviews based on supervised machine learning techniques (SMLT).The analysis of dataset by supervised machine learning technique (SMLT) to capture several information’s like, variable identification, univariate analysis, bi-variate and multi-variate analysis, missing value treatments and analyze the data validation, data cleaning/preparing and data visualization will be done on the entire given dataset. To propose a machine learning-based method to classify the Google play store reviews results in the form of positive, neutral or negative best accuracy from comparing supervise classification machine learning algorithms. Millions of mobile apps are available in app stores, such as Apple’s App Store and Google Play. For a mobile app it would be increasingly challenging to stand out from the enormous competitors and become prevalent among users. Good user experience and well-designed functionalities are the keys to a successful app. To achieve this, popular apps usually schedule their updates frequently. If we can capture the critical app issues faced by users in a timely and accurate manner, developers can make timely updates, and good user experience can be ensured.

Enhancement of Short Term Prediction Using modified LSTM
Author's Name: Gayathri G S, Vimala Devi A, Priya B, Kalaiarasi C

Abstract— In today’s economy there is a profound impact of the stock market. The main aim of stock market prediction is to predict the future values of the financial stocks of a company. Predicting the stock market is the toughest task in the field of computations. The recent trend in stock market prediction uses the machine learning technology which makes predictions based on the values of current stock market indices by training on their previous values. Stock market prices depend very much on demand and supply. High demand stocks will increase in price while heavy selling stocks will decrease. Fluctuations in stock prices affect investor perception and thus there is a need to predict future share prices and to predict stock market. The historical stock prices include opening, closing, volume weighted average price, turnover, trade values are more helpful in estimating the future of a stock. In existing method stock prices are predicted by using several machine learning algorithms like Linear Regression, Long short term memory (LSTM), Autoregressive integrated moving average (ARIMA). This project proposes modified LSTM algorithm to get better results for short term prediction of stock prices and may yield increase in accuracy.

An implementation overview of deep learning in medical imaging focusing on MRI
Author's Name: M.Raju, R.Tanusree, S. Rajukumar

Abstract— What has happened in machine learning lately, and what does it mean for the future of medical image analysis? Machine learning has witnessed a tremendous amount of attention over the last few years. The current boom started around 2009 when so-called deep artificial neural networks began outperforming other established models on a number of important benchmarks. Deep neural networks are now the state-of-the-art machine learning models across a variety of areas, from image analysis to natural language processing, and widely deployed in academia and industry. These developments have a huge potential for medical imaging technology, medical data analysis, medical diagnostics and healthcare in general, slowly being realized. We provide a short overview of recent advances and some associated challenges in machine learning applied to medical image processing and image analysis.

IOT based Health Machine for covid-19 with Matlab
Author's Name: Varalakshmi K, Dharma Prakash V, Noble Lourdhu, Deepika H

Abstract— The novel Corona Virus (COVID 19) is a pandemic of unimaginable proportion and magnitude that is posing a great challenge worldwide to the medical industry in the 21st century. It has completely changed the texture of life to a greater extent. IoT is considered as one of the fastest growing technology in the medical and industrial fields. The proposed device can help the patients by checking their health status through monitoring and diagnosing patients. This paper presents a measuring and recording device for heart rate, body temperature and CT imaging. An Arduino device with sensors will be used to measure these records and send them to the cloud server.

Intelligent Voice Over System for Disabled People
Author's Name: Gayathri G S, Vimala Devi A, Priya B, Kalaiarasi C

Abstract— The problem of multimodal interaction is discussed. The use of blinking and winking, interpreted as Eye gestures, is considered. The main aim of this study is to propose a simple method that allows the recognition of the state of the eye: open or closed; and to distinguish between blinking and winking. A Few people and groups are not able to operate the computer because of their illnesses. In this scenario, it is sound, to introduce a method of computer operation, which is easily accessible, even considering the disabilities of the differently abled. A Human-Computer Interaction system that is designed for individuals with severe disabilities to simulate control of a traditional computer mouse is introduced. A specific human computer interaction system using eyeball movement is presented. But in this system, we use eyes instead of mouse which provides a unique way of operating the computer with the help of eyeball movements. This system tracks the eye movements of the user with arduino data and simulates the eye movements into mouse cursor movements on screen and also detects the user's eye staring at the icon and will translate it into click operation on screen. Embedded c to read sensor value to controller, and control pushes the data to computer via serial communication. Python reads serial data from USB and converts signal data to UI interface with pre-trained data and conditions. The main aim of this system is to help the user to control the cursor without the use of hands and is of great use especially for the people with disability.

improved Levenberg–Marquardt neural network for energy efficiency and anomaly detection in WSN
Author's Name: S.Manjunath, R.Mahinaram and M.Deepa

Abstract— One of the key goals in the design of the networks is to increase the lifespan of wireless sensor networks (WSNs). Using different models of intelligent energy management could help designers achiseve this objective. By reducing the number of sensors required to collect data on the environment, these models can achieve higher levels of energy efficiency without sacrificing the quality of the readings. When battery power is an issue, wireless sensor networks (WSNs) are often employed for applications such as monitoring or tracking. Several routing protocols have been developed in the last several years as possible answers to this problem. Despite this, the issue of extending the lifetime of the network while considering the capacities of the sensors remain open. As a result of applying neural networks, Low-Energy Adaptive Clustering Hierarchy (LEACH) and Energy-Efficient Sensor Routing (EESR) can be improved in terms of their overall efficiency as well as their level of dependability, as is shown in this research EESR.

Image Processing Based on Optical Character Recognition with Text to Speech for Visually Impaired
Author's Name: Vijayanarayanan A, Savithiri R, Lekha P, Abbirami R S.

Abstract— The present paper has introduced an innovative, efficient and real-time cost beneficial technique that enables users to hear the contents of text images instead of reading through them. It combines the concept of Optical Character Recognition (OCR) and Text to Speech Synthesizer (TTS) using a webcam. The major problem faced by visually impaired people these days is that they are unable to do text recognition on their own, which forces them to depend on others for their day to day activities such as reading newspapers, letters sent through post, referring books etc. The ultimate aim of the project is to help the visually impaired people to recognize the text. When a printed text is shown in front of the webcam it has to capture the image, extract the text from the image and should read out the text either through computer audio or headphones. Text-to-Speech (TTS) is the ability of a computer to produce spoken words by converting text to voice. In other words Text-to-Speech software is a speech synthesizer that vocalizes text in real time in a natural way. This paper describes the design, implementation and experimental results of the device. This device consists of two modules, image processing module and voice processing module.

An implementation of Intelligent Street Light Managing System using LoRa Network
Author's Name: M.Raju, R.Tanusree, S. Rajukumar

Abstract— Smart meters are used in smart buildings to regulate streetlights (SLs) that enhance implementation performance by making evaluations of electricity parameters and illumination levels easily accessible. Temporary high-resolution data is needed in places where smart meters are being used in order to increase SLs' energy effectiveness (EE). Because of its low expenditure, secure connections, and extended range both inside and outside, long range (LoRa) is a perfect wireless communication technology for usage in smart urban. To do this, a low-cost new scheme has been developed and thoroughly tested by creating three devices based on the Arduino open-source electrical framework: the Control device for Street Lights (CSL), Lighting Level Device (LLD), and Gateway LoRa Network (GLN). The creation and execution of the hardware and software are discussed in this work. The energy efficiency for street lights (ESL) methodology, that also uses the illumination level evaluated on a single set of SLs with such a hyperparameter tuning, has also been established to conservation and efficiency of lighting systems.

Real-Time Facial Recognition Based Student Proctoring System Using KNN Algorithm
Author's Name: Varalakshmi K, Dharma Prakash V, Noble Lourdhu, Deepika H

Abstract— Individual uniqueness and individuality are their faces. In this project, human faces are automatically used for attendance tracking purpose. Student attendance is essential for any college, university, and school. The traditional way to record attendance is to call the student’s name or registration number and the attendance will be recorded. The time required for this purpose is a major concern. Assume that the course will take approximately 60 minutes and registration for attendance will take around 10 minutes. For any teacher, this is a waste of time. To avoid these losses, this project uses an automated process based on image processing. This project uses face detection and face recognition. Face detection is used to locate the face area, and face recognition is used to mark the presence of the understudy. KNN algorithm is used for obtaining the machine learning model alias KNN Classifier. A classifier of all students in the class is saved, and attendance is recorded if each student’s face matches one of the faces in the trained model.

Segmentation of Spatial and Geometric Information from Floorplans using CNN Model
Author's Name: Sheik Rahman, Raja Mohammed, M. Padmadevi

Abstract— In automated document analysis, floorplans are a concern for several years and algorithmic approaches have been used until recently. This problem has also improved output with the emergence of Convolutionary neural networks (CNN). In this study, it is the task to retrieve space and geometric data from planes as accurately as possible and the bulk of the information is extracted from a plane image by means of instance segments, such as the Cascade Mask R-CNN. A new way of using keypoint CNN is presented in order to supplement the segmentation, so that precision corner positions can be identified. Then they are coupled to the resulting segmentation in a post-processing stage. With an average IoU of 72.7% compared to 57.5%, the resulting segmentation scores surpass the existing benchmark for CubiCasa5k floorplan datasets.

Support Vector Machine based Multi Block Chain E Voting System
Author's Name: Gayathri G S, Vimala Devi A, Priya B, Kalaiarasi C

Abstract— In India, the voting system generally uses the manual approach where voters queue up in a physical space to cast their votes for their choices. Manual voting system without any doubt does not lead to 100% voting rate. This project primarily focuses to provide the people an authentic scientific community with parties’ policy positions, with respect to their constituencies. A web application is developed using reactJS that aims to analyze the role of the existing political parties/independents who are contesting in the forthcoming state election. This web application also helps the state government in achieving 100% voting rate in the state elections by providing an e-voting system enabling Face Detection authentication. The project is mainly aimed at providing a secured and user-friendly Online Voting System. The problem of voting is still critical in terms of safety and security. This system deals with the design and development of a web-based voting system using face detection and Aadhaar card in order to provide a high performance with high security to the voting system. The proposed Online Voting System allows the voters to scan their faces, which is then matched with an already saved image within a database that is retrieved while casting of votes. The voting system is managed in a simpler way as all the users must login by Aadhaar card number and click on his/her favorable candidates to cast the vote by using face detection it provides enough security which reduces the dummy votes

Mechanical Behaviour of Jute / Glass / Chicken Feather Reinforced Hybrid Composites
Author's Name: Sathiyamoorthy Margabandu , Madhavan. P Anilkumar, Prabhakaran, Sivakumar

Abstract— The hybridization of natural and synthetic fibers leads to composites optimum mechanical properties. Chicken feather fibers have potential application in light weight composites. In this study, an attempt was made to study the effect of the stacking sequence on epoxy-based Jute (J), Glass (G), Chicken feather (CF) hybrid composites. Two types of hybrid composite, each containing two different layers of jute and glass fabric and chicken feather as filler were manufactured by the hand layup method. Mechanical properties, such as tensile and flexural were studied. The experimental results showed that the stacking sequence of the fiber layers has a significant effect on the overall performance of prepared hybrid composites. Among the prepared hybrid composites, composite laminate with jute layers on their outer surfaces exhibited improved tensile strength of 92.3 MPa and the composites with glass fiber layers on their outer surfaces showed enhanced flexural strength of 125.7 MPa. This study demonstrated a new way to utilize CFFs and the prepared hybrid fiber composite has better mechanical properties than CFF composites while having increased bio-based content.

Step up – Step down Converters Based on Semi-active Rectifiers for High Output Voltage Applications
Author's Name: S.Dinakar Raj, S.L. Sreedevi, R.Tamilamuthan, Dr.G.Irin Loretta, A.VijayaLakshmi, P.Yamunaa

Abstract— A systematic method for developing isolated step up – step down (IBB) converters is proposed in this paper, and single-stage power conversion, soft-switching operation and high efficiency performance can be achieved with the proposed family of converters. On the basis of a non-isolated two-switch step up - step downconverter, the proposed IBB converters are generated by replacing the DC step down-cell and step up-cell in the non-isolated step up - step down converter with the AC step down - cell and step up - cell, respectively. Furthermore, a family of semi-active rectifiers (SARs) is proposed to serve as the secondary rectification circuit for the IBB converters, which helps to extend the converter voltage gain and reduce the voltage stresses on the devices in the rectification circuit. Hence, the efficiency is improved by employing a transformer with a smaller turns ratio and reduced parasitic parameters, by using low-voltage rating MOSFETs and diodes with better switching and conduction performances. A full-bridge IBB converter is proposed and analyzed in detail as an example. The phase-shift modulation strategy is applied to the full-bridge converter to achieve isolated step up - step down conversion. Moreover, soft-switching performance of all active switches and diodes can be achieved over a wide load and voltage range by the proposed converter and control strategy. A 380V output prototype is fabricated to verify the effectiveness of the proposed family of converters,

Multipurpose File Transfer and File Inquiry
Author's Name: Varalakshmi K, Dharma Prakash V, Noble Lourdhu, Deepika H

Abstract— A Company wants to sync their files based on the configuration. Also inquire files based on Metadata. The Preconditions are the API will transfer files across local, FTP, sFTP locations. A REST API will be able to retrieve documents based on metadata. The post conditions are the data files should be transferred according to the requirement. The Success case would be the data files should be transferred according to the requirement, the Documents can be retrieved based on the metadata. The high volume PDF file migration is achieved by converting the PDF files to Base64.