Dr.ANBARASAN M
@tagore-engg.ac.in
Department of Information Technology
Tagore Engineering College
Scopus Publications
Scopus Publications
M. Anbarasan, S. Prakash, A. Antonidoss, and M. Anand
Springer Science and Business Media LLC
Sudheer Reddy Bandi, M Anbarasan, and G Merlin Linda
IEEE
Synthetic Aperture Radar (SAR) and Optical sensors are very popular depending on their strengths and application needs. The ocean surface will be covered with clouds most of the time and optical sensors cannot capture the information as they cannot operate in dust and cloudy conditions. So, SAR sensors come into existence to capture the atmospheric phenomena in those regions. There is a wide range of texture extraction techniques such as traditional-based wavelet analysis, statistical analysis, local pixel-based approaches, and dimensional reduction approaches. However, these approaches have limitations in boundary extraction, high-dimensional feature extraction, and uniformity representation. In recent times, deep learning (DL) approaches to extract features which varies based on the applications, image interpretation, and so on. At the same time, DL approaches are limited in transparency, computation requirements, and hyperparameter tuning. In contrast, the proposed S5ELBP i.e., Supervised SAR image classification using the integration of Support Vector Machine (SVM) and Scale Selective Extended Local Binary Pattern (SSELBP) extraction techniques shall overcome various challenges of existing traditional and DL approaches such as adaptive thresholding, less computational resources, and noise robustness. Hence, the objective of the proposed methodology is to extract the image features from the Scale Selective Extended Local Binary Pattern (SSELBP). Also, the SSELBP features are integrated with SVM to show the strength of the proposed approach in SAR image classification. Additionally, the performance of the present work is enhanced by comparing the extracted results with various classifiers. Finally, the work is concluded by recommending the scope of future work.
D Meenakshi, M Anbarasan, S Murugesan, and B Selvalakshmi
IEEE
Deep learning (DL), an evolving technology, allows computers to learn autonomously from historical data. Furthermore, it employs a number of methodologies to generate the precise models and create predictions based on facts or old data. A deep learning system produces the predictions based on mathematical models of prediction which is built from past information once it gets new information. Nowadays, the use of deep learning in medical systems is crucial in various applications such as recognition, segmentation, and classification. Worldwide, lung illness is a prevalent occurrence in human beings which includes Covid-19, pneumonia, emphysema, asthma, lung cancer, etc. The early identification of lung diseases helps to detect the disease more accurately. Several deep learning techniques, including convolutional neural network, vanilla neural networks, visual geometry group-based networks and Caps Net, are used to predict the lung illness. To forecast different lung diseases from chest X-ray and CT scan pictures, typical CNN’s performance and learning framework is insufficient. This proposed work aims to build a lung disease detection model using CNN with various hyperparameter tuning to enhance the accuracy of the model. Also, the accuracy is associated with advanced methods to enhance the performance of the proposed work.
M. Prabha, M. Anbarasan, S. Sunithamani, and Mrs. K. Saranya
IEEE
In recent years, the wireless sensor networks are widely used in numerous real-time applications such as WBAN monitoring and tracking. Recent developments in wireless networks have given rise to new and reliable methods for enhancing network lifetime, energy efficiency, and scalability. The power consumption of the entire wireless sensor network and the energy level of each individual sensor node are closely related by the commonly used clustering technique to manage the sensor networks. In order to optimize data transmission, the fuzzy C means algorithm is employed in this article to thoroughly analyze the cluster head while considering the energy that is available in each node and distance metrics from the base station. This study demonstrates how carefully choosing cluster heads and node clustering, which divides large networks into smaller clusters, can enhance the lifespan of a network. The proposed network uses a multi-hop routing approach, where each sensor node can independently collect and send data in order to address the data rate issue. The suggested cluster routing protocol was put to the test with 1000 data transmission rounds to demonstrate its strengths and weaknesses in terms of network lifetime and energy efficiency. The choice of the cluster head node, the distance between the nodes, and the amount of energy needed for subsequent data transmission are all considered to be random for each round. The simulation results show that the suggested methodology beats cutting-edge routing techniques and achieves a promising network performance. Furthermore, the effects of hierarchical cluster head selection point to the potential of our method for use in WSN in the future. The following tests were performed using computer simulation, including comparing the effect of network life on the increased number of rounds before and after the influence of an energy-efficient routing protocol, and examining performance metrics for network lifetime.
Wang Yue, Changgang Yu, A. Antonidoss, and M Anbarasan
American Society for Photogrammetry and Remote Sensing
In modern society, energy conservation is an important consideration for sustainability. The availability of energy-efficient infrastructures and utilities depend on the sustainability of smart cities. The big streaming data generated and collected by smart building devices and systems contain useful information that needs to be used to make timely action and better decisions. The ultimate objective of these procedures is to enhance the city's sustainability and livability. The replacement of decades-old infrastructures, such as underground wiring, steam pipes, transportation tunnels, and high-speed Internet installation, is already a major problem for major urban regions. There are still certain regions in big cities where broadband wireless service is not available. The decision support system is recently acquiring increasing attention in the smart city context. In this article, a deep learning–based sustainable decision support system (DLSDSS) has been proposed for energy building in smart cities. This study proposes the integration of the Internet of Things into smart buildings for energy management, utilizing deep learning methods for sensor information decision making. Building a socially advanced environment aims to enhance city services and urban administration for residents in smart cities using remote sensing techniques. The proposed deep learning methods classify buildings based on energy efficiency. Data gathered from the sensor network to plan smart cities' development include a deep learning algorithm's structural assembly of data. The deep learning algorithm provides decision makers with a model for the big data stream. The numerical results show that the proposed method reduces energy consumption and enhances sensor data accuracy by 97.67% with better decision making in planning smart infrastructures and services. The experimental outcome of the DLSDSS enhances accuracy (97.67%), time complexity (98.7%), data distribution rate (97.1%), energy consumption rate (98.2%), load shedding ratio (95.8%), and energy efficiency (95.4%).
Ying Wang, BalaAnand Muthu, and M. Anbarasan
World Scientific Pub Co Pte Ltd
In recent studies, YOLOv3, a deep learning-based target detection algorithm, becomes extensively used in object recognition, especially guiding the visually disabled. Current YOLOv3-based assistive technology for the disabled person can now achieve high-precision, real-time object recognition. Even though this algorithm has several flaws, including the failure to estimate distances and the difficulty of accurately recognizing points in fog or haze, it can perform well in waste management. Therefore, this study proposes an Intelligent Garbage Monitoring Scheme based on an improved YOLOv3 Target Detection Algorithm (IGMS-iYTDA) to classify the IoT’sgarbages (IoT) enabled trash can. The performance of the proposed scheme has been evaluated and illustrated for various classification evaluation metrics. The evaluation results show the highest classification accuracy of 99.9% compared to existing models for the proposed scheme.
Humin Yang, M. Anbarasan, and Thanjai Vadivel
World Scientific Pub Co Pte Ltd
Artificial intelligence can open modern opportunities and potentials for smart education. Smart learning purposes at providing holistic learning to learners utilizing modern technologies to fully prepare them for a fast-evolving world where adaptability is vital. With the advancement of technologies and within modern society, smart education will pose several challenges, like educational structures, pedagogical theory, educational ideology, technology leadership, and teachers’ learning leadership. Therefore, in this paper, an Intelligent Knowledge-based recommender system (IKRS) has been proposed using artificial intelligence for smart education. The recommendation is generated by the genetic algorithm and K-nearest neighbor algorithm (KNN) utilizing the optimized weight attributes vectors that signify the learner’s opinions. The experimental results show that the suggested IKRS model enhances student-teacher interaction, student involvement level, learning quality and predicts students’ learning style compared to other existing methods.
Balaji N, Lakshmi S, Anand M, Anbarasan M, and Mathiyalagan P
Wiley
Sudheer Reddy Bandi, M. Anbarasan, and D. Sheela
Czech Technical University in Prague - Central Library
Sensors of different wavelengths in remote sensing field capture data. Each and every sensor has its own capabilities and limitations. Synthetic aperture radar (SAR) collects data that has a high spatial and radiometric resolution. The optical remote sensors capture images with good spectral information. Fused images from these sensors will have high information when implemented with a better algorithm resulting in the proper collection of data to predict weather forecasting, soil exploration, and crop classification. This work encompasses a fusion of optical and radar data of Sentinel series satellites using a deep learning-based convolutional neural network (CNN). The three-fold work of the image fusion approach is performed in CNN as layered architecture covering the image transform in the convolutional layer, followed by the activity level measurement in the max pooling layer. Finally, the decision-making is performed in the fully connected layer. The objective of the work is to show that the proposed deep learning-based CNN fusion approach overcomes some of the difficulties in the traditional image fusion approaches. To show the performance of the CNN-based image fusion, a good number of image quality assessment metrics are analyzed. The consequences demonstrate that the integration of spatial and spectral information is numerically evident in the output image and has high robustness. Finally, the objective assessment results outperform the state-of-the-art fusion methodologies.
Zhuomin Sun, M. Anbarasan, and D. Praveen Kumar
Wiley
N. Balaji, S. Lakshmi, M. Anand, M. Anbarasan, and P. Mathiyalagan
Springer Science and Business Media LLC
M. Anbarasan, S. Prakash, A. Antonidoss, and M. Anand
Springer Science and Business Media LLC
MANET(Mobile Adhoc Networks) possess the open system condition, absence of central server, mobile nodes that make helpless to security assault while conventional security components couldn’t meet MANET security prerequisites in view of restricted correspondence data transfer capacity, calculation power, memory and battery limit in addition to the vitality enabled environment. The trusted MANETs provide a reliable path and efficient communication but the secrecy of the trust values sometimes may be overheard by the masqueraders. Due to the need of the clustered MANETs the exchange of mathematical values remains to be a necessary part. In the proposed security of the trusted MANETs is focused so as to provide rigid and robust networks when additional resources are added. For clustering of the nodes LEACH protocol is suggested in which the CHs and CMs are fixed for the data transfer in the network. The energy is disseminated in the LEACH as to avoid the battery drain and network fatal. Hence to add resistance and to make an authentic network, the encryption and decoding is incorporated as a further supplementary to avoid the denial of service attacks, we have utilized DoS Pliancy Algorithm in which the acknowledgment based flooding attacks is focused. Likewise the encoded messages from the source node in one cluster can be recoded in the transmission stage itself to reproduce the messages. Contrasted with the past works, QoS of our proposed work has been made strides when tested with black hole and sink hole attacks. Simulation results shows that the DoS pliancy scheme works better and efficient when compared to the existing trust based systems.
M. Anbarasan, BalaAnand Muthu, C.B. Sivaparthipan, Revathi Sundarasekar, Seifedine Kadry, Sujatha Krishnamoorthy, Dinesh Jackson Samuel R., and A. Antony Dasel
Elsevier BV
Abstract Natural disasters could be defined as a blend of natural risks and vulnerabilities. Each year, natural as well as human-instigated disasters, bring about infrastructural damages, distresses, revenue losses, injuries in addition to huge death roll. Researchers around the globe are trying to find a unique solution to gather, store and analyse Big Data (BD) in order to predict results related to flood based prediction system. This paper has proposed the ideas and methods for the detection of flood disaster based on IoT, BD, and convolutional deep neural network (CDNN) to overcome such difficulties. First, the input data is taken from the flood BD. Next, the repeated data are reduced by using HDFS map-reduce (). After removal of repeated data, the data are pre-processed using missing value imputation and normalization function. Then, centred on the pre-processed data, the rule is generated by using a combination of attributes method. At the last stage, the generated rules are provided as the input to the CDNN classifier which classifies them as a) chances for the occurrence of flood and b) no chances for the occurrence of a flood. The outcomes obtained from the proposed CDNN method is compared parameters like Sensitivity, Specificity, Accuracy, Precision, Recall and F-score. Moreover, when the outcomes is compared other existing algorithms like Artificial Neural Network (ANN) & Deep Learning Neural Network (DNN), the proposed system gives is very accurate result than other methods.
M. Anand, K. Thinakaran, N. Kalyana Sundaram, and M. Anbarasan
Elsevier BV
Abstract This paper presents a new strategy for choosing an ideal way between the hubs for information transmission in portable impromptu organization for industrial automation on improving communication between sensors for defining materials. Portable impromptu organization is a brief, framework less organization that conveys in multi-jump way. Directing is a significant issue in versatile impromptu organizations. Regularly, directing convention may chooses the way between two hubs dependent on numerous measurements like separation between the hubs, bounces check, delay, trust esteems, leftover vitality, gotten signal quality and so forth. In this proposed strategy, two courses are found between the source and objective and afterward by considering the separation metric ideal course is chosen for information transmission between the hubs. Reenactment results show that the proposed Multiple Metric based Twofold Route Selection AODV Routing protocol (MMBTRS_AODV) beats the ordinary AODV directing convention. The proposed steering convention brings about high bundle conveyance proportion with less normal start to finish deferral and low directing overhead.
Anbarasan M., Prakash S., Anand M., and Antonidoss A.
Wiley
Vitality reliant and mobility are the two factors that deplete some measure of energy during the radio communication often occur in Mobile Ad hoc Network (MANET) among and between the nodes. Since MANET is a decentralized system, it poses to the exploitation of energy during the reliable path selection during routing. Hence, selection of protocol plays a major aspect in MANET for combatting against the energy depletion of nodes, Instable link, mobility character, and load balancing. The RPS‐LEACH protocol is proposed in this paper for the efficient path selection by organizing the nodes into clusters. For each cluster, Cluster Head (CH) is elected based on its strong transmission power and battery life. The Cluster Members (CM) is organized, respectively, based on its ability to respond to its cluster head REQs. Mainly, RPS values are calculated by the two parameters such as (i) successful interactions and (ii) unsuccessful interactions. In the proposed system, RPS (Reliable Path Selection) values are exchanged among the nodes for the reliable path selection by choosing the highest RPS value node for its communication. Hence, by adopting the RPS scheme, the nodes are able to choose the most reliable path for the efficient data transfer by exchanging the RPS values by ignoring the paths that consuming long time to transfer the data. In this procedure, correspondence convention, which comprises imperative impact of vitality liberality of such networks, is addressed. In cases of applications with no predefined or intended infrastructure, this scheme paves way for the versatile hubs to perform as the energy efficient routing nodes. Simulation carried out in NS2 and results shown improved performance when compared with LDTS and RPS‐LEACH.
Anand M., Sasikala T., and Anbarasan M.
Wiley
Mobile ad‐hoc network (MANET) is a gathering of portable nodes that works without foundation or central administration. Because of the accessibility of little and cheap remote conveying nodes, MANETs can be utilized as a part of different applications, for example, front line correspondence and debacle alleviation applications. Energy consumption is an important issues in MANET because the mobile nodes are battery powered, hence diminishing system lifetime as batteries get depleted rapidly as nodes move and change their positions quickly crosswise over MANET. We propose an energy efficient channel aware routing algorithm for mobile ad‐hoc networks, called energy efficient channel aware ad‐hoc on‐demand multipath distance vector routing (EECA‐AOMDV). EECA‐AOMDV addresses three vital prerequisites of mobile ad‐hoc networks: energy effectiveness, unwavering quality and dragging out system lifetime. The proposed energy efficient channel mindful AOMDV (EECA‐AOMDV) utilizes the channel normal nonfading span and nodal residual energy as directing metric to choose the stable route for way revelation. The key thought of the convention is to discover average channel nonfading duration and maximum nodal residual energy as routing metrics of each course during the time spent choosing way and sort the multi course by slipping channel nonfading duration and nodal leftover energy.
M. Anbarasan, S. J. S. Prasad, R. Meenakumari, and P. A. Balakrishnan
IEEE
The control of temperature in the different zones of barrel heating system in plastic molding machine is a challenging task, as it requires quick rise time without overshoot. The conventional PID controller does not produce the desired performance for this application. This paper ends in presenting the idea of using PI-PD controller-a modified structure of PID controller for controlling the temperature of the melt in the barrel. Fuzzy logic technique is used for obtaining the optimum values of the controller parameters. The controller is modeled in SIMULINK. From the simulation results, it is observed that the fuzzy tuned PI-PD controller outperforms the other controllers namely Ziegler-Nichols tuned PID controller, Fuzzy Logic controller and Fuzzy tuned PID controller.