Browse Items (2150 total)
Sort by:
-
Enhancing the Recognition of Hand Written Telugu Characters: Natural Language Processing and Machine Learning Approach
Handwritten character recognition has wider application in many areas including heritage documents, education, document digitalization, language processing, and assisting the visually handicapped and other related areas. The paper tries to improve the accuracy and efficiency of recognizing handwritten letters of Telugu language scripts, a difficult task for computers. Telugu is most widely spoken language in southern part of India, it has rich cultural heritage. Using the Natural Language Toolkit (NLTK), this study investigates ways to enhance recognition accuracy by analyzing handwritten content and implementing methods such as feature extraction and classification. The purpose is to use NLTK's capabilities to develop handwritten character recognition. 2024 IEEE. -
Real-World Application of Machine Learning and Deep Learning
The world today is running on the latest computer technologies and one of those is machine learning. The real life example that most of us know is speech recognition. Google Assistant is the common example for this Speech recognition. This google assistant is not only limited till 'Ok Google', but it responds to all your questions in a smart way. It can manage all your calls or can book appointments. Imagine you fell down while de-boarding a bus. So, Next time you take care so that you don't fall that is something that your brain has interpreted from your past experience. This is what exactly deep learning is, it imitates human brain works. Deep learning is sub-branch of machine learning. It is able to build all new things based on its previous experiences. Many of us have heard about driverless cars and medical diagnosis. Recently google has developed a new technology where all your cardiovascular events can be predicted by eye scan so, that doctors can get a clear view of what is inside the body of a patient. These all are developed using machine learning. It has a capability to change the human world into a complete robotic world. Anyways, it also has its own disadvantages. This article discusses about those, Scope of machine learning, its Market potential, financial growth and Current applications of machine learning. 2019 IEEE. -
Preprocessing Big Data using Partitioning Method for Efficient Analysis
Big data collection is the process of gathering unprocessed and unstructured data from disparate sources. As data deluge, the large volume of data collected and integrated consist missing values, outliers, and redundant records. This makes the big dataset insignificant for processing and mining knowledge. Also, it unnecessarily consumes large amount of valuable storage for storing redundant data and meaningless data. The result obtained after applying mining techniques in this insignificant data lead to wrong inferences. This makes it inevitable to preprocess data in order to store and process big dataset effectively and draw correct inferences. When data is preprocessed before analytics the storage consumption is less and computation and communication complexity is reduced. The analytics result is of high quality and the needed time for processing is considerably reduced. Preprocessing data is inevitable for applying any analytics algorithm to obtain valuable pattern. The quality of knowledge mined from large volume of big data depends on the quality of input data used for processing. The major steps in big data preprocessing include data integration from disparate sources, missing value imputation, outlier detection and treatment, and handling redundant data. The process of integration includes steps such as extraction, transformation, and loading. The data extraction step gathers useful data used for analytics and the transformation process organize the collected data in structured format suitable for analytics. The role of load process is to store transformed data into secured storage so that data can be obtained and processed effectively in future. This work provides preprocessing techniques for big data that deals with missing values and outliers and results in obtaining quality data partitions. 2023 IEEE. -
Investigation on thermal barrier effects of 8YPSZ coatings on Al-Si alloy and validation through simulation
In high temperature engineering field, protection of metal components operating at high temperatures has been a problem since the attempts to realize high efficiency aero engines in the 1940s. Researchers have been working on finding a solution for this issue and thermally insulating the surface of the base metal component with a suitable high temperature material, generally a ceramic, is one solution. The Thermal Barrier Coatings, popular worldwide as TBCs have found wide spread applications in aerospace and automobile industry after its successful application in aerospace engines in mid 1970s. In the field of aerospace, generally a super alloy will be the substrate and in automobile field this process is very much suited on aluminium casting alloys, which is the raw material for high speed diesel engine cylinder blocks and pistons. Although a good quantity of research work on TBCs have been completed in the field of aerospace, the published literature on such coatings on Aluminium castings alloys are limited. Present research aims to throw some light in this grey area by plasma spray coating Aluminium-Silicon (Al-Si) substrates with popular Yttria Partially Stabilized Zirconia as top coat and underlying nickel aluminide bond coat. Al-Si alloys are widely used in automobiles. Experiments were conducted to evaluate the temperature drop across a 250 mm thick TBC at different ceramic surface temperatures and then validating the experimental results by simulation in ANSYS. Experimental results and simulated results showed a close match, thereby validating the findings. 2019 Elsevier Ltd. All rights reserved. -
Residual stresses analysis on thermal barrier coatingsndt tool for condition assessment
Improvement in the engine efficiency follows reduction in fuel consumption which is possible by increasing the engine combustion temperature. Coating the piston of diesel engine with a high temperature-resistant material, known as thermal barrier coating, generally 68% Y2O3 stabilized ZrO2, is a popular method to reduce the temperature it experiences in service and to increase engine efficiency. Whether bare or coated component, fabrication and different thermal expansion coefficients of the ceramic coating and piston metal cause generation of residual stresses in them. These hidden residual stresses (tensile or compressive) play a significant role in governing the failure mechanism of the different sections of the components and their important role (also developed in service) is mostly neglected in engineering practices. Residual stresses analysis of components in service may throw light on the condition of the components without destroying them. In this work, portable X-ray residual stress analyzer was used to evaluate the condition of AlSi alloys piston flat plates that were coated with 250-m-thick 68% Y2O3 stabilized ZrO2 and subjected to thermal treatments. The analysis revealed (a) residual stress-free pattern for uncoated AlSi substrate, (b) compressive residual stress at the substrate (AlSi)coating (TBC) interface and (c) tensile residual stress at the substrate (AlSi)coating (TBC) interface of a thermal shocked coated substrate. The analysis method exhibited good possibility for using this as a tool for non-destructive testing for predicting the onset of failure at the coating substrate interface, without destroying the component in service. Springer Nature Singapore Pte Ltd 2020. -
Protection offered by thermal barrier coatings to Al-Si alloys at high temperatures - A microstructural investigation
Thermal barrier coatings, with ~50 mm thick Nickel-Aluminide bond coat and ~250 mm thick Yttria-Stabilized zirconia ceramic top coats were synthesized by Air Plasma Spray coating process on flat plates machined from Al-11Si alloy diesel engine pistons. Coating process parameters and qualifications that were followed were based on previous studies made on the same substrates. The ceramic coatings were subjected to various thermal treatments such as (a) thermal shock cycling tests and (b) continuous heating in a furnace. Uncoated Al-Si samples were simultaneously subjected to the same thermal treatments and used as reference to study the protection offered by the coatings to the base metal substrates. Thermal shock cycles tests involved subjecting the coated and uncoated Al-Si plates to oxy-acetylene flame to allow the ceramic surface to be maintained at 500 C for 1000 cycles (one cycle comprised of heating for 60 s, withdrawal from flame and forced cooling in ambient air for 60 s) and similar thermal shock cycles in an electric furnace. The specimen were also heated in a furnace at 300 C for 1000 continuous hours. Stresses induced during thermal shock cycles and oxidation of bond coat-ceramic coat interface during the exposure to heat are the main reasons for the coating's failure. Details of an investigation on the microstructural changes and oxidation behaviour of the substrate and the ability of the coatings to protect the metal substrates from oxidation are presented. Microstructural studies were carried out by employing a Scanning Electron Microscope attached with Energy Dispersive X-ray spectroscopy facility. The findings were compared on (a) uncoated Al-Si alloy and (b) thermal barrier coated Al-Si alloy with a goal to understand the capability of the coatings to protect the metal from the influences of thermal treatments, at temperatures lower than the melting point of the Al-Si alloy. 2019 Elsevier Ltd. All rights reserved. -
Thermal Barrier Coating Development on Automobile Piston Material (Al-Si alloy), Numerical Analysis and Validation
This work is focused on the thermal barrier coating (TBC) development on aluminium-silicon (Al-Si) alloy casting materials, widely used as automobile components (cylinder blocks, pistons etc.). TBCs enable enhanced combustion within the chambers of diesel engines resulting in improved performance and components life. Uniform coating thickness development on complex contours of automobile pistons is a challenging task worldwide which results in varying thermal barrier characteristics across the non-uniform thickness. In consistent (in thickness) coatings are most likely to lead to uneven thermal barrier effects across the TBC thicknesses which directly affect the performance and the lubrication system of the engine. This warrants the development of stable and consistently thick coatings for ideal performance of the Low Heat rejection (LHR) engine. The present research work involved building different thicknesses (100, 125 and 150?m) of commercial 6-8%Yttria stabilized zirconia (YSZ) TBCs on 50? to 75? thick nickel aluminide (NiAl) bond coat. The influence of thickness on thermal barrier characteristics via experimentation and numerical analysis has been studied. Flat plates machined from automobile pistons were used as substrates. The coatings were characterized for thermal barrier effects for hot ceramic surface face temperatures up to 550C (by using oxy-acetylene flame to heat up the TBC surface), structural phase analysis by X-ray Diffraction (XRD) and microstructure analysis in metallographic cross section by employing Scanning Electron Microscope (SEM). An analytical investigation also was carried out to determine the approximate temperature at each interface. A code was developed to calculate the temperature drops across the coated plate and the net heat available at the coated surface using MATLAB. This is important considering the effects, small changes in temperatures will bring on the creep life on the metal. 2019 Elsevier Ltd. -
Assessment of thermal barrier effects across 8%Y2O3-ZrO2 coatings on Al-Si substrates via electrical heating source
Ceramic Thermal Barrier Coatings (TBCs) provide protection to metals from degradation at high temperature. A major factor deciding the effectiveness of the coating in service is the temperature drop across the thickness of the coating. Common practice to determine the temperature drop is to subject the coating with a high heat providing flame with preset velocity by using combustible gases focused towards the coated surface, that keep the surface at desired stabilization temperature and the temperature is measured at the back side of the coating, i.e. at the metal side. The challenge is to heat the complete specimen surface using the flame and to reach an accurate stabilization temperature by using the flame as the heating source. In the present work, this challenge was overcome by using a uniform source of heat i.e. an electric heater on the entire coating surface. This paper presents the results obtained by studying the thermal barrier effects across TBCs by using the electrical heating source that provided the heat on the ceramic surface in a controlled and uniform manner, thereby establishing a newer assessment method. The TBCs were prepared by plasma spray coating commercial 8%Yttria-Stabilized Zirconia (8YSZ) as the top ceramic coat on flat plates of Aluminium 11% Silicon alloy removed from diesel engine pistons. TBC thicknesses varied between 100?m and 600?m. Air Plasma Spray coating was employed to coat the substrates which initially were spray coated with 50-75 ?m thick bond coat of Nickel Aluminide. Thermal barrier test was conducted by heating the entire coated surface uniformly and by keeping the ceramic surface temperature constant till the stabilization in the range of 300C to 500C. The temperature drop achieved was in the range of 46C to 127C depending upon the coating thickness. Details of the tests conducted and results obtained are presented. 2019 Author(s). -
Challenges in Plasma Spraying of 8%Y2O3-ZrO2 Thermal Barrier Coatings on Al Alloy Automotive Piston and Influence of Vibration and Thermal Fatigue on Coating Characteristics
Although Thermal Barrier Coatings (TBCs) have found extensive application in automotive engines to enhance performance and to reduce fuel consumption and pollution, challenges of obtaining uniform and consistent coatings on non-uniform and irregularly shaped components are overcome only when the coatings are deposited via robot controlled APS or EBPVD. Atmospheric Plasma Spraying (APS) is the most commonly used and relatively cost-effective method to make TBCs: but not all APS facilities are equipped with comprehensive coating accessories. In a reciprocating diesel engine, the bowl at the piston crown forms one side of the combustion chamber and includes the space between piston crown (generally 9% Si-Al alloy in light - medium duty diesel fuel vehicle) and cylinder head. To achieve maximum effective fuel spray distribution and combustion, normally the crown of the piston has complex contours. One of the many service related parameters to be monitored to reduce the innumerable faults contributing to the performance of the engine is vibration. This paper addresses the issue related with the challenges associated with the plasma spraying of consistent and adherent TBC on Al-9% Si research pistons and its complex contours by APS, subjecting the coated pistons to thermal fatigue tests and evaluation of the coating characteristics after subjecting to vibration. 2018 Elsevier Ltd. -
Advanced Approaches for Hate Speech Detection: A Machine and Deep Learning Investigation
The prevalence of online social media platforms has led to an alarming rise in the frequency of cyberbullying and hate speech. This study uses a variety of machine-learning approaches and deep- learning algorithms to identify hate speech. The goal is to create a thorough and successful method for locating and categorizing hate speech on online networks. Our suggested approach intends to deliver a comprehensive solution to address the urgent problem of cyberbullying and hate speech in the digital sphere by leveraging the strength of these cutting-edge techniques. We work to make social media users' online experiences safer and more welcoming by identifying and addressing such harmful online actions. Through rigorous experimentation, we evaluate the efficacy of these methodologies, ultimately revealing that the Bidirectional Gated Recurrent Unit (Bi-GRU) outperforms the other employed techniques. The Bi-GRU model demonstrates superior hate speech detection capabilities, substantiated by robust performance metrics. This research contributes to the field by providing empirical evidence that deep learning models, such as Bi-GRU, can significantly advance hate speech detection accuracy. The findings underscore the potential of leveraging advanced neural architectures in the pursuit of fostering a more inclusive and respectful digital space. 2024 IEEE. -
Sentiment Analysis on Indian Government Schemes Using Twitter data
People use social media for entertainment, fetching information, news, business, communication and many more. Few of such social media applications are Facebook, Twitter, WhatsApp, Snapchat and so on. Twitter is one among the micro blogging websites. We are using Twitter mainly because it has gained a lot of media attention. The text written is referred to as tweets, where a common man can tweet or can write their hearts out. We would be fetching the direct responses from the public and hence the data is more real-time. First step is to fetch the tweets on a particular scheme using python language code followed by the cleaning process then comes the creation of bag of words. Later these bags of words are given as an input to the algorithms. Finally, after training the algorithms, we will be getting the sentiment of the public on that scheme. 2019 IEEE. -
Enhanced Horse Optimization Algorithm Based Intelligent Query Optimization in Crowdsourcing Systems
Crowdsourcing is a strategy of collecting information and knowledge from an abundant range of individuals over the Internet in order to solve cognitive or intelligence intensive challenges. Query optimization is the process of yielding an optimized query based upon the cost and latency for a given location based query. In this view, this article introduces an Enhanced Horse Optimization Algorithm based Intelligent Query Optimization in Crowdsourcing Systems (EHOA-IQOCSS) model. The presented EHOA-IQOCSS model mainly based on the enhanced version of HOA using chaotic concepts. The proposed model plans to accomplish a better trade-off between latency and cost in the query optimization process along with answer quality. The EHOA-IQOCSS is used to compute the Location-Based Services (LBS) namely K-Nearest Neighbor (KNN) and range queries, where the Space and Point of Interest (POI) can be obtained by the conviction level computation. The comparative study stated the betterment of the EHOA-IQOCSS model over recent methods. 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG. -
An Efficient Fuzzy Logic Cluster Formation Protocol for Data Aggregation and Data Reporting in Cluster-Based Mobile Crowdsourcing
Crowdsourcing is a procedure of outsourcing the data to an abundant range of individual workers rather than considering an exclusive entity or a company. It has made various types of chances for some difficult issues by utilizing human knowledge. To acquire a worldwide optimal task assignment scheme, the platform usually needs to collect location information of all workers. During this procedure, there is a major security concern; i.e., the platform may not be trustworthy, and so, it brings about a threat to workers location privacy. Recently, many distinguished research papers are published to address the security and privacy issues in mobile crowdsourcing. According to our knowledge, the security issues that occur in terms of data reporting were not addressed. Secure and efficient data aggregation and data reporting are the critical issue in Mobile Crowdsourcing (MCS). Cluster-based mobile crowdsourcing (CMCS) is the efficient way for data aggregation and data reporting. In this paper, we propose a novel procedure, the efficient fuzzy logic cluster formation protocol (EFLCFP) for cluster formation, and use cluster cranium (CC) for data aggregation and data reporting. We recommend a couple of secure and efficient data transmission (SET) protocols for CMCS, (i) SET-IBE uses additively homomorphic identity-based encryption system and (ii) SET-IBOOS uses the identity-based online/offline digital signature system, respectively. Then, we have widen the features of cluster cranium by increasing the propensity to achieve aggregation and reporting on the data yielded by the requesters without scarifying their privacy. Also, considering query optimization using cost and latency. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Grading of Red Chilli, Cardamom and Coriander Using Image Processing
Indian cuisine is known for its wide range of spices. Spices are known as the heart and soul of Indian food. Traditionally, categories are identified based on certain chemical technology or with the help of senses gifted to mankind. In this paper, an image processing technique used to extract multiple features is presented to determine the various categories of spices consumed. This proposed work uses different varieties of common Indian spices such as Capsicum annuum (dry red chilli), Elettaria cardamomum (cardamom) and Coriandrum Sativum (coriander). While creating the image dataset, different categories of all spices were taken from southern region of India. Features are extracted from the manually created image dataset, which forms the base for classification. The result obtained using Multilayer Perceptron (MLP), Naive Bayes and Random Forest classifier is found to be optimal. 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. -
Growth and characterization of chalcogenide crystals by vapour method
A horizontal linear gradient two zone furnace was designed and employed to grow single crystals of indium telluride by Physical Vapour Deposition (PVD) method. It was calibrated for various trials including, series and parallel combinations of coils, and set temperatures. Systematic growth runs for chalcogenide crystals were performed by varying the source and growth temperatures. Crystals of different sizes and morphologies were obtained. The morphology and chemical analysis of the grown crystals were investigated by Scanning Electron Microscope (SEM) and Energy Dispersive Analysis using X-rays (EDAX). The hardness of the crystals was estimated using a Vickers microhardness tester. 2011 American Institute of Physics. -
Multimodal Face and Ear Recognition Using Feature Level and Score Level Fusion Approach
Recent years have seen a significant increase in attention in multimodal biometric systems for personal identification especially in unconstrained environments. This paper presents a multimodal recognition system by combining feature level fusion of ear and profile face images. Multimodal biometric systems by combining face and ear can be used in an extensive range of applications because we can capture both the biometrics in a non-intrusive manner. Local texture feature descriptor, BSIF is used to extract discriminative features from biometric templates. Feature level and score level fusion is experimented to improve the performance of the system. Experimental results on different public datasets like GTAV, FEI, etc., show that the proposed method gives better performance in recognition results than individual modality. The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024. -
Ear Recognition Using Pretrained Convolutional Neural Networks
Ear biometrics, which involves the identification of a person from an ear image, is challenging under unconstrained image capturing scenarios. Studies in Ear biometrics reported that the Convolutional Neural Network is a better alternative to classical machine learning with handcrafted features. Two major concerns in CNN are the requirement of enormous computing resources and large datasets for training. The pretrained network concept helps to use CNN with smaller datasets and is less demanding on hardware. In this paper, three pre-trained CNN models, AlexNet, VGG16, and ResNet50 are used for ear recognition. The fully connected classification layers of the nets are trained with AWE, an unconstrained ear dataset. Alternatively, the CNN layers output (the CNN features) are extracted, and an SVM classification model is built. To improve the classification accuracy, the training dataset size is increased through data augmentation. Data augmentation improved the classification accuracy drastically. The results show that ResNet50, with the fully connected classification layer, results in higher accuracy. 2021, Springer Nature Switzerland AG. -
A Novel Approach to Automatic Ear Detection Using Banana Wavelets and Circular Hough Transform
Ear is an attractive biometric trait that maintain their structure with increasing age. Because of the complex geometry of ear, its detection is very difficult. This paper proposes a modified algorithm for automatic detection of 2D ear images using Banana wavelets and Hough transform. Banana wavelets derived from bank of stretched and curved Gabor wavelets are used to identify curvilinear ear structure. Addition of a preprocessing stage, prior to application of banana wavelets is found to improve the detection results further. The proposed algorithm is brought in to comparison with three existing algorithms and evaluated on standard databases. In addition to manual detection accuracy, this paper also calculates the efficiency of the proposed method using automatic classification techniques. The features like LBP and Gabor extracted from segmented ear image is used by different classifiers to determine whether the segmented portion of the image is class Ear or Non ear. 2019 IEEE. -
Breaking News Recognition Using OCR
Identifying and recognition of breaking news in most of the TV channels in different backgrounds with varying positions from a static image plays a significant role in journalism and multimedia image processing. Now a days its very challenging to isolate only breaking news from headlines due to overlapping of many categories of news, keeping all this in mind, a novel methodology is proposed in this paper for detecting specific text as a breaking news from a given multimedia image. Basic digital image processing techniques are used to detect text from the images. The methods like MSER (Maximally Stable Extremal Regions) and SWT (Stroke Width Transform) are used for text detection. The proposed work focuses on extraction of text in breaking news images also discusses the different methods to overcome existing challenges in text detection along with different types of breaking news datasets collected from various news channels are used to identify text from images and comparative study of different text detection methods. The comparative study proves that MSER and SWT is a better technique to detect text in images. Finally using OCR (Optical Character Recognition) technique to extract the breaking news text from the detected regions will help in easy indexing and analysis for journalism and common people. Extensive experiments are carried out to demonstrate the effectiveness of the proposed approach. 2019, Springer Nature Singapore Pte Ltd. -
Neural Network based Student Grade Prediction Model
Student final grade GPA is the collective efforts of their previous and ongoing efforts of each semester examination may predict accurately using the neural network which receives the input weight of each matrix element of variables to next neuron. The GPA prediction based on regular class performance and previous grades with background variables were found much significant. This research tries to explore the model comparison and evaluate student grade prediction using various neural network models. The single-layer half i.e., successful student model predicts 90 total accuracies than the single layer with five hidden layer neurons (88.5 percent). The multi-layer with two hidden layers (7,3) is 84 percent accuracy is less than one percent accuracy than multilayer with three hidden layers. Similarly, the multilayered with four hidden layered 25,12,7,3 model predicts the least accuracy (77 percent accuracy) for student grade. Similarly, the passed student prediction model has less accuracy than both students' 86 percent. 2022 IEEE.