Ebook: Smart Intelligent Computing and Communication Technology
Recent developments in the fields of intelligent computing and communication have paved the way for the handling of current and upcoming problems and brought about significant technological advancements.
This book presents the proceedings of IConIC 2021, the 4th International Conference on Intelligent Computing, held on 26 and 27 March 2021 in Chennai, India. The principle objective of the annual IConIC conference is to provide an international scientific forum where participants can exchange innovative ideas in relevant fields and interact in depth through discussion with their peer group. The theme of the 2021 conference and this book is ‘Smart Intelligent Computing and Communication Technology’, and the 109 papers included here focus on the technological innovations and trendsetting initiatives in medicine, industry, education and security that are improving and optimizing business and technical processes and enabling inclusive growth. The papers are grouped under 2 headings: Evolution of Computing Intelligence; and Computing and Communication, and cover a broad range of intelligent-computing research and applications. The book provides an overview of the cutting-edge developments and emerging areas of study in the technological fields of intelligent computing, and will be of interest to researchers and practitioners from both academia and industry.
The book presents high-quality research papers presented in the 4th International conference, IConIC 2021, organized by Panimalar Engineering College, Chennai, Tamil Nadu, India on 26th and 27th March, 2021. The conference proceedings have a complete track record of the papers that are reviewed and presented at the conference. The principle objective is to provide an international scientific forum wherein the participants can mutually exchange their innovative ideas in relevant fields and interact in depth through discussion with peer groups. Both inward research as well as core areas of Intelligent Computing and its applications will be covered during these events. The aim of the proceedings is to provide cutting-edge developments taking place in the technological fields of Intelligent Computing which will assist the researchers and practitioners from both academia as well as industry to exchange, cross-fertilize their ideas and update knowledge in the latest developments and in the emerging areas of study. Researchers are now working in the relevant areas and the proceedings of IConIC 2021 plays a major role to accumulate those significant works in a single arena. The theme of the book ‘Smart Intelligent Computing and Communication Technology’ will focus on technological innovation and trendsetting initiatives applicable for corporate, industries, education, security to improve the business value, optimize business processes and enable inclusive growth, Proven IT governance, standards and practices that has led to the development in the form of prototype, design and tools to enable rapid information flow to the user. The book is divided into two parts namely: “Evolution of Computing Intelligence” and “Computing and Communication”.
The above parts will bring together the ideas, innovations and the experimental results of academicians, researchers and scientists in their domain of interest, in the areas of Intelligent Computing, Communication, and Control systems. The recent developments in the intelligent Computing and Communication paves the way for handling the current and upcoming problems thereby bringing a drastic change in the technological advancements. Furthermore, the newer and innovative ideas will be well groomed with adequate technical support and the core competent technical domains in Computing Intelligent Systems and Communication to improve the intellectual aspects. The conference looks for significant contributions towards the latest technologies in theoretical and practical aspects. Authors are solicited to contribute their experiences that describe significant advances in the following areas, but are not limited. It also provides a premier platform for Scientists, Researchers, Practitioners and Academicians to present and discuss about the most recent innovations, trends and concerns as well as practical challenges encountered in this field.
The Book Editors
Prof. V.D.Ambeth Kumar & Prof. Dr. S.Malathi
Department of Computer Science and Engineering, Panimalar Engineering College, Chennai, Tamil Nadu, India
Prof. Valentina Emilia Balas,
Department of Computer Science, “Aurel Vlaicu” University of Arad, Arad, Romania
Prof. Margarita Favorskaya,
Department of Information Systems and Technologies, Reshetnev Siberian State University of Science and Technology, Krasnoyarsk, Russia
Prof. Dr. Thinagaran Perumal,
Universiti Putra Malaysia, Malaysia
Image compression is the processing of images by using Transform operations and Encoding techniques. Nowadays, there is an essential need of these methods in the medical field. This work is focused on the performance quality assessment of medical images using Image Compression Techniques. Image compression is the process of compression of an image in such a way that it has less space than the original image. It is an organization of compression technique that reduces the size of an image file without affecting its quality to a greater extent. The bio-orthogonal transform is used for decomposing the Lung images. After decomposition, different methods of encoding are performed and finally the proposed compression methods are evaluated for finding optimum algorithm for medical image.
An attempt has been made to develop an algorithm for banks to check the credibility of borrowers to avoid nonperformance assets. People move towards different banks for loan purpose to fulfil their financial needs. Approaching bank for loan is increasing day by day mainly for child marriage, education, agriculture, business, home loan etc. Some people take the loan and they won’t pay back in time or some will move out of the country without any intimation, so that bank will go in loss. Even now in covid-19 pandemic many industries were closed but they need to give salary to the employees, need to pay rent and electricity bills too for that they will approach bank for loan. For all these cases bank first need to analyse their Credit Information Bureau India Limited score and check whether they had done loan repayments in appropriate time or not. In the present work the effectiveness of K nearest neighbor algorithm were analysed. This research were carried out using python. The accuracy of this classifier is analysed using following metrics such as Jaccard index, F1-score and LogLoss. This helps to find the potential of the customer which is much higher than the data mining classification algorithm and thus it helps in sanctioning loans.
Recently machine learning algorithms are utilized for identifying network threats. Threats otherwise called as intrusions, will harm the network in a stern manner, thus it must be dealt cautiously. In the proposed research work, a deep learning model has been applied to recognize and categorize unanticipated and unpredictable cyber-attacks. The UNSW NB-15 dataset has a vital number of features which will be learned by the hidden layers present in the suggested model and classified by the output layer. The suitable quantity of layers, neurons in each layer and the optimizer utilized in the proposed work are obtained through a sequence of trial and error experiments. The concluding model acquired can be utilized for estimating future malicious attacks. There are several data preprocessing techniques available at our disposal. We used two types of techniques in our experiment: 1) Log transformation, MinMaxScaling and factorize technique; and 2) Z-score encoding and dummy encoding technique. In general, the selection of data preprocessing techniques has a direct impact on the output performed by any machine learning process and our research, attempts to prove this concept.
Sickle cell anemia is a blood disorder which is widespread across world with an estimate that about 30% of total population will be affected by the end of 2050. Especially in India, the frequency of sickle cell trait can reach up to 35%, thereby increasing attention over the topic for research. It is necessary to detect the disorder as much as it is for finding a cure to it. Therefore, a system that detect the sickle cells from the blood smear images or erythrocytes images is required. The affected cells change their shape from circular to sickle shape, due to the presence of typical hemoglobin called hemoglobin S. Nowadays, the elongated affected cells are identified using image processing techniques. The shape descriptors which are the vital features used to identify any shapes in an image. The contrast and brightness of blood smear images may not be consistent as the acquisition of these images depends on various factors like luminance and chrominance. To detect the sickle cells properly in the blood smear images, the histogram equalization technique is applied to improve the contrast of the image. The image is converted to binary in order to find the boundary of the cells in the image. In the binary image, each normal cell will have holes, Since the normal red blood cell have the shape of a doughnut that has been pressed in the middle slightly. The boundary of all the cells in the image is traced and the holes inside each cells are filled. Each cell is considered as a connected component for which the eccentricity is calculated. The eccentricity is the shape descriptor which is the ratio of the major axis and minor axis of the connected components. If the eccentricity is greater than threshold, the cells are identified as sickle cells.
Every person has a unique personality. A set of characteristics or qualities possessed by an individual is called as the personality traits. Interpersonal dependency refers to the amount of support that an individual gets from the other person and the level of it gets varied among the people. Those who have higher interpersonal dependency are labeled as low self- esteemed person. The present study is initiated to identify the traits possessed by an interpersonal dependent. An adapted questionnaire is used for the study. Totally the questionnaire consists of seventeen questions among which ten questions measured the variable personality traits and the remaining seven questions measured the variable interpersonal dependence. The data was collected from fifty employees working in corporate for the present study. Content and face validities are used to assess the validity. Inferences are drawn through the use of descriptive and inferential statistic.
Leaf diseases in cucumber must be identified early and accurately as it reduces the yield. Disease detection in plants is the important field of Artificial Intelligent in agriculture. This causes a periodic outbreak of plant diseases which leads to crop death. Automatic plant disease detection may be benefits in monitoring large scale cultivation of crops, and detects the disease symptoms at early stage itself. Here to detect the disease present in the cucumber the tensor flow technique is used and then artificial neural network is used to classify these diseases. The common disease like cucumber mosaic virus, leaf miner, bacterial leaf spot, and leaf blight are examined with disease detection accuracy of 98.66%. This image processing techniques and algorithm are designed using python language to segment the diseases and categories them whether infected or healthy with the help of Pycharm software.
Lane Detection is the fundamental vehicle driving system and self-driving. The proposed concept is to employ the pixel difference in the expected lane and the road backdrop to detach the lane from the road, and then the Curve fitting model is made use of in the segregated lanes to locate the straight line in the image as the lane line. This paper offers a lane detection method based on the Sobel filter and Curve-fitting model for lane line tracking in different conditions. The main objective is to improve the accuracy of the Xi’an city database and the KITTI vision benchmark suite dataset. To achieve this HLS color space was performed which identifies the lane by adding pixel values.
In today’s day of Modern era when the data handling objectives are getting bigger and bigger with respect to volume, learning and inferring knowledge from complex data becomes the utmost problem. Almost all of the real-world information are maintained under a relational fashion holding multiple relations unlike orthodox approaches containing single relational as a whole. Moreover several fields viz. biological informatics, microbiology, chemical computations needed some more dependable and expressive approach which can provide more sophisticated results with faster speed. Hence in context with multi-relational data mining in which data is directly retrieved from different records without dumping into single table, we have described a novel approach of improved Multi-Relational Decision Tree Learning Algorithm based on the implementations. In this paper provided a comparative study of the aforementioned approach in which we have taken certain results from the literature review. Experiments mainly includes results from widely used datasets viz. Mutagenesis database which demonstrates that Multi-Relational Decision Tree Learning Algorithm provides a promising alternative from previous conventional approaches such as Progol, FOIL, and Tilde.
Web based Social Network (SN) is one of the popular mediums where the information is shared among users. The combined data from different sources available on the open web needs to be evaluated for its quality and trustworthiness. A trustable community can be formed, if the social networking applications, media, forums, etc., can automatically judge the user who requests to join the community. It is necessary to know the trustability of the persons within the community to share the official and personal information. Therefore, a trust model has become an essential part of Online Social Network (OSN) to find the trustworthiness of the requester. Significance of the work detailed here is to model a trustworthy user suggestion in OSN applications.
In search engines, though feature-based query is provided, Content Based Image Retrieval (CBIR) still results in less sensitivity and specificity. It is because the conventional approach is based on feature extraction and inherent parameters in conventional feed forward networks. Performance of the system is strongly dependent on the extracted features. Hence it is necessary to develop a CBIR system that retrieves the similar images without explicit feature extraction. Convolutional Neural Networks are the recent neural network architectures which accept images as input and perform both feature extraction and classification. The proposed work aims at using the conventional architectures of VGGNET, RESNET, and DENSENET for flower classification in CBIR. Performance is measured in terms of accuracy of classification.
In the current world millions of people are suffering with bone diseases such as osteoporosis and osteopenia. The early detection of osteoporosis and osteopenia disease is very important as it helps people to be cautious and get treated on time. Hence research on early detection of osteoporosis and osteopenia disease has gained importance across the world. In this paper, analysis of suitable kernel for Support Vector Machine (SVM) focussing on the classification of osteoporosis and osteopenia disease has been carried out and presented. The kernel functions considered includes polynomial, linear, RBF and Gaussian to find the optimal one for the classification of osteoporosis and osteopenia disease with improved accuracy.
The rising usage of smartphones and sensing models in today’s life induces the development of large scale urban sensing networks in communication concepts. With the efficient implementation of people centric mobility models, the personal communication devices of people are acting as the sensor nodes that are capable of sensing the human behaviours and participating in framing the smart protocol designs for smart cities. Without the involvement of mobile users or people, the new protocols for mobility management, traffic management, environmental sensing and other applications become futile. The proposed cloud based model improves the reliability and scalability of the system with its multiple cloud servers design. The single point of failure can be resisted since many cloud servers belonging to a provider is used. So the framework remains fault tolerant in the presence of any server attacks. A standard homomorphic based encryption scheme is used for providing data confidentiality and also data is transferred anonymously improving the privacy of the system. The data aggregation process is supported by the model protecting the user’s privacy. The performance analysis for the proposed framework is done in terms of design goal analysis and computation cost analysis.
This paper proposes the design of the prosthetic arm by reconstructing the structure and proportions of an amputated arm using high precision methods and dimensions. To achieve this, CT images of the patient’s amputated and non-amputated arm are collected from the Rehabilitation centre. The patient CT data were imported to a 3D modelling software i.e., Mimics Innovation Suite version 22.0 Materialise 3-Matic version 14.0 original licensed software. The exported file is given to the Computer-Aided Design software, the geometry of the socket and the prosthetic arm were designed according to the mirrored geometry of the non-affected arm. 3D rendering for various degrees of movement has been carried out for animation.
The Blockchain based I-Voting uses a digital-currency analogy where in eligible voters can cast a ballot anonymously using a computing environment. BIV(Blockchain I-Voting) employs an encrypted key, smart biometrics and tamper proof real-time personal ID verification. Blockchain enable the creation of tamper-proof audit trails for voting. In this generation of Technology there is an effect way for casting vote through online(Digital Voting) to make a public electoral process faster, cheaper, and more easier. In this generation it is being a compelling one in modern society which removes a power barrier between the elected candidate and the voter.
Segmentation of Text, the undertaking of partitioning a record obsessed by adjoining sections dependent on its semantic design, is a longstanding test in language understanding. Each segment has its applicable significance. Those segments arranged as phrase, text group, point, express or any data unit relying upon the errand of the content examination. This paper proposes the profound learning-based content segmentation strategies in NLP where the content has been portioned utilizing quick tangled neural organization. We propose a bidirectional LSTM prototype where text group embedding is gotten the hang of utilizing fast RNNs and the phrases are fragmented dependent on context-oriented data. This prototype can consequently deal with variable measured setting data and present an enormous new dataset for text segmentation that is naturally divided. Besides, we build up a segmentation prototype dependent on this dataset and show that it sums up well to inconspicuous regular content. We find that albeit the segmentation precision of FRNN with Bi-LSTM segmentation is advanced than some other segmentation techniques. In the proposed framework, every content is resized obsessed by required size, which is straightforwardly exposed to preparation. That is, each resized text has foreordained and these phrases are taken as fragmented content for preparing the neural organization. The outcomes show that the proposed framework yields great segmentation rates which are practically identical to that of segmentation-based plans for manually written content.
In the previous era, a computer is programmed for some specific task. An electronic device is programmed to do its function electronically. It was done with a target device, the programming environment and the system. We get the necessary intermediate code by running the program with the above said environment and committed into the target device. Thus the device performs the task it was intended to do. In case if we need to change the functionality of the device by the learning experience of the vendor and users, the vendor will upgrade the product. Nowadays in this machine learning era, the devices are programmed in such a way it can learn by its own experience and with the available data it collected it can even manipulate the algorithm by itself with the provided data set. Thus machine learning is ruling this era. We are going to discuss the machine learning algorithms here which was used to predict by itself with the data set collected. Therefore, machine learning is all about learning about computer algorithms that progress its potential through the experience. Thus, Machine learning is presently highly regarded analysis topic and applied to all told application in day to day life. In this paper we have a tendency to extract the knowledge of machine learning algorithms like decision tree, Naive Bayes and enforce the algorithms with sample dataset of weather prognostication.
Video annotation technique delivers many additional video processing capabilities for several applications. Sports broadcast video content is unique in regard to wealth of information as compared to any other video. Sports video annotation is becoming popular among researchers in recent times because of wide range of applications and challenges it pose. The demand for optimized design of framework for sports video annotation is at peak. This paper surveys state-of-the-art in annotation framework design, particularly for sports applications and provides insight into future aspects. This survey may help researchers to further conceive and develop advanced universal frameworks applied to all sports.
The most shocking events were was the recent discovery of the fraudulent activities in the Punjab National Bank. This is due to frequent systemic failures that detect human errors. Blockchain technology is the greatest solution for this issue. It is surprisingly common for the information settlement mechanism like SWIFT to be on a isolated ledger from the payment settlement mechanism. If the banks uses a ledger that stores information settlement distributed across all the participants, then the fraudulent user may reflect on all the available participants in the transactions, auditors and regulators. Our Paper is a Decentralised Loan Management Web Application (DApp) built on Ethereum blockchain which targets on preventing such fraudulent attacks on Loans sanctions by decentralising the processes. The security features authentication of the user identity, authentication of bank officials and multiple levels of verification of details are implemented using Public Key Infrastructure (PKI).
Archeological assets of the nation are to be preserved and rejuvenated. Ageing of these sites poses a major challenge in assessing the health of these structures. Hence it necessitates a technique that is non contact non invasive and non hazardous. Passive InfraRed Thermography is one such technique that uses an IR camera to capture the temperature variations. Thermal variations are mapped as thermographs. Interpretation of thermographs provides information about the health of the archeological structures. As the paradigm has shifted to computer aided interpretation, segmentation techniques and line profiling are used for describing the hotspot. Of the various segmentation techniques, morphological image processing provides accurate segmentation of cold spot.
The COVID-19 pandemic has become an evolving situation all around the globe and the spread is at an alarming rate. Adapting public health-informed hygiene practices can control the transmission of COVID-19, basic measure being wearing a mask. Face detection is the process of identifying faces in a given input image or video and indicating it by drawing a bounding box around the face. In this paper we introduce a deep learning computer vision model to recognise if a person visible through the camera is wearing a mask or not. The deep learning algorithm administered in our work is a transfer learning based Convolutional Neural Network (CNN) and the face detection task is done using the Viola-Jones algorithm approach.
Technological transformation is unlocking new opportunities in wearable devices used in sports application. Nowadays training the sports involves the use of integrating smart sensors, cameras, internet of things and intelligent data algorithms into a device which is wearable making the players to achieve their maximum performance. These smart devices replace the coach and manage all aspects of technical training except for the physical training given by the real coach. This paper provides a comprehensive study on the intelligent data analysis made on the data acquired from sensors to give a meaningful sense to it. The smart training methods employed currently in various sports are identified and presented. The future directions in this area of research are also presented.
In day-to-day life cancer is the severe disease to cure. This paper deals about the cervical cancer and its related issues. The cervical cancer is a preventable disease, but it is one of the second most leading causes of cancer deaths in women. There are two main types of cervical cancer namely squamous cell carcinoma, adenocarcinoma. The major stages of cervical cancer will be dealt by evaluating with clear diagrams and also the major risk factors of cervical cancer are discussed. Finally the various treatment methods of cervical cancer are used to analyze and present the related issues.
Food recognition system is essential framework in modern world, the motivations behind this recognition systems are assessing calorie and nutrients of food varieties and show the formula of the food. As all the processing on picture recognition is performed on a personal digital assistant (PDA), the framework doesn’t have to send pictures to a server and runs solution to a normal advanced mobile phone in a constant one. Spontaneous image-based food acknowledgment is an isolated inspiring task. Conventional picture analysis preparing methodologies have fallen off with low order classification prediction in the past, while deep learning methodologies permitted the recognizable proof of food things. To recognize food items, a client ought to need to draw bounding boxes by contacting the screen first, and afterward the framework starts food thing acknowledgment inside the showed bounding boxes. Moreover, the framework assesses the direction of food areas with the higher SVM (Support Vector System) yield score is expected upon to be acquired, show it as an arrow on the screen to request that a client move an advanced mobile phone camera. This acknowledgment cycle is performed on and on about once each second. We will carry out this task as an Android advanced mobile phone application to utilize numerous CPU cores effectively for modern real-time food recognition system.