KUMARESAN PERUMAL

Prof. KUMARESAN PERUMAL

Vellore Institute of Technology, Vellore
orcid

Research Focus & Metrics

Academic Career Length

13Years (2014-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 September 2020 in Sensors
Verified
Journal Article
Open access
PDF file
Chiranji Lal ChowdharyChiranji Lal Chowdhary
Chiranji Lal Chowdhary
PPPushpam Virenbhai Patel
Pushpam Virenbhai Patel
KKKrupal Jaysukhbhai Kathrotia
Krupal Jaysukhbhai Kathrotia
MAMuhammad Attique
Muhammad Attique
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
MIMuhammad Fazal Ijaz
Muhammad Fazal Ijaz
The majority of imaging techniques use symmetric and asymmetric cryptography algorithms to encrypt digital media. Most of the research works contributed in the literature focus primarily on the Advanced Encryption Standard (AES) algorithm for encryption and decryption. This paper propose an analysis for performing image encryption and decryption by hybridization of Elliptic Curve Cryptography (ECC) with Hill Cipher (HC), ECC with Advanced Encryption Standard (AES) and ElGamal with Double Playfair Cipher (DPC). This analysis is based on the following parameters: (i) Encryption and decryption time, (ii) entropy of encrypted image, (iii) loss in intensity of the decrypted image, (iv) Peak Signal to Noise Ratio (PSNR), (v) Number of Pixels Change Rate (NPCR), and (vi) Unified Average Changing Intensity (UACI). The hybrid process involves the speed and ease of implementation from symmetric algorithms, as well as improved security from asymmetric algorithms. ECC and ElGamal cryptosystems provide asymmetric key cryptography, while HC, AES, and DPC are symmetric key algorithms. ECC with AES are perfect for remote or private communications with smaller image sizes based on the amount of time needed for encryption and decryption. The metric measurement with test cases finds that ECC and HC have a good overall solution for image encryption.

13 July 2020 in Sensors
Verified
Journal Article
Open access
PDF file
Chiranji Lal ChowdharyChiranji Lal Chowdhary
Chiranji Lal Chowdhary
Mohit MittalMohit Mittal
Mohit Mittal
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
PPP. A. Pattanaik
P. A. Pattanaik
Zbigniew MarszałekZbigniew Marszałek
Zbigniew Marszałek
The herpesvirus, polyomavirus, papillomavirus, and retrovirus families are associated with breast cancer. More effort is needed to assess the role of these viruses in the detection and diagnosis of breast cancer cases in women. The aim of this paper is to propose an efficient segmentation and classification system in the Mammography Image Analysis Society (MIAS) images of medical images. Segmentation became challenging for medical images because they are not illuminated in the correct way. The role of segmentation is essential in concern with detecting syndromes in human. This research work is on the segmentation of medical images based on intuitionistic possibilistic fuzzy c-mean (IPFCM) clustering. Intuitionist fuzzy c-mean (IFCM) and possibilistic fuzzy c-mean (PFCM) algorithms are hybridised to deal with problems of fuzzy c-mean. The introduced clustering methodology, in this article, retains the positive points of PFCM which helps to overcome the problem of the coincident clusters, thus the noise and less sensitivity to the outlier. The IPFCM improves the fundamentals of fuzzy c-mean by using intuitionist fuzzy sets. For the clustering of mammogram images for breast cancer detector of abnormal images, IPFCM technique has been applied. The proposed method has been compared with other available fuzzy clustering methods to prove the efficacy of the proposed approach. We compared support vector machine (SVM), decision tree (DT), rough set data analysis (RSDA) and Fuzzy-SVM classification algorithms for achieving an optimal classification result. The outcomes of the studies show that the proposed approach is highly effective with clustering and also with classification of breast cancer. The performance average segmentation accuracy for MIAS images with different noise level 5%, 7% and 9% of IPFCM is 91.25%, 87.50% and 85.30% accordingly. The average classification accuracy rates of the methods (Otsu, Fuzzy c-mean, IFCM, PFCM and IPFCM) for Fuzzy-SVM are 79.69%, 92.19%, 93.13%, 95.00%, and 98.85%, respectively.

1 March 2025 in Environmental Research Communications
Verified
Journal Article
Open access
RVRamesh V
Ramesh V
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
Predicting crop yields remains a crucial challenge in agriculture, as these forecasts influence decision-making at global, regional, and individual crop levels. Historically, such predictions have utilized diverse data sources, including agricultural, land, climatic, atmospheric, and other pertinent information. Over the past several years, the application of machine learning techniques has emerged as a valuable analytical approach for estimating agricultural productivity, thereby informing decisions regarding crop selection and management strategies throughout the entire growing cycle. Various kinds of machine learning models have been utilized in research to forecast crop yields. Our work proposes a stacked ensemble model designed for the purpose of predicting crop yield. The proposed model employs a stacked ensemble learning approach, with a Decision Tree Regressor functioning as the meta-model to amalgamate predictions from six distinct base learner models: Linear Regression (LR), Elastic Net, XGBoost Regressor, K-Neighbors Regressor (KNR), AdaBoost Regressor, and Random Forest Regressor (RFR). The proposed stacked ensemble model achieves superior crop yield prediction performance, evidenced by a notable enhancement in accuracy and a significant decrease in RMSE, surpassing the predictive capabilities of traditional machine learning models. The ensemble model’s performance was assessed using several metrics, including a Mean Absolute Error of 7.20 tons/hectare, Mean Square Error of 15570.32 tons2/hectare2, Root Mean Square Error of 124.78 tons/hectare, and Coefficient of Determination (R2 Score) of 0.98. The performance results demonstrate that stacked ensemble model outperforms other conventional machine learning approaches, achieving a high R-squared score of 98%.

9 December 2024 in IEEE Access
Verified
Journal Article
Open access
GSG. Sunilkumar
G. Sunilkumar
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
Cardiovascular disorders are the primary cause of death on a global scale. The World Health Organization report indicates that approximately 18 million people die from CVD each year. Major cardiac risks include arrhythmia and coronary artery disease, among others. Recent advancements in Artificial Intelligence play a pivotal role for life-saving interventions in CVD treatment. This survey examines the latest progress in Machine Learning, Deep Learning and Pre-trained transfer learning models for classifying and predicting CVD using a survey of 159 articles, including 36 image datasets, 41 signal data and 52 clinical data from various sources. The survey investigates cardiac risk factors, cardiac dysfunction classification, various modalities and medical image processing techniques, performance metrics and hybrid techniques. Surveys on traditional neural networks such as Convolutional Neural Networks, Artificial Neural Networks, and Recurrent Neural Networks often achieve an accuracy rate of 70% to 95%. Leveraging pre-trained architectures such as ResNet, DenseNet, AlexNet, MobileNet, EfficientNet and GoogLeNet, transfer learning models consistently outperform other approaches frequently achieving accuracy levels greater than 96%. Researchers utilize various hybrid optimization algorithms to improve overall accuracy rate. The outcome of the survey supports a precise prognosis for patients with comorbidities. The survey findings indicate limitations in the incorporation of multimodal data for real-time risk assessment. The study results possess the capacity to bridge gaps in research on cardiovascular disease prediction, thereby assisting medical practitioners in early identification and subsequent prognosis.

1 February 2020
Verified
Conference Paper
Closed access
RGRamya Govindaraj
Ramya Govindaraj
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
KHK. Sree Harshitha
K. Sree Harshitha
In this research work. Voting is commonly related to politics and is finished with often exploitation and manual approach where voters stand to vote for his or her decisions. Manual voting may lead to malpractices sometimes.so there is a need to implement online voting system. This is for expand the technology from manual voting system to digital voting system. In this specific research our idea is to implement online voting system with features like the schemes that the specific party has implemented, based on the features we are going to vote. The main reason we need to shift from normal voting system to online voting system is that we can consume our time and can vote from anywhere through online. We have implemented this by using C# as a programming language, Microsoft SQL server 2012 and Microsoft azure as a cloud.

21 February 2024 in Diagnostics
Verified
Journal Article
Open access
PDF file
SRSivashankari Rajadurai
Sivashankari Rajadurai
KUMARESAN PERUMALKUMARESAN PERUMAL
KUMARESAN PERUMAL
MIMuhammad Fazal Ijaz
Muhammad Fazal Ijaz
Chiranji Lal ChowdharyChiranji Lal Chowdhary
Chiranji Lal Chowdhary
Malignant lymphoma, which impacts the lymphatic system, presents diverse challenges in accurate diagnosis due to its varied subtypes—chronic lymphocytic leukemia (CLL), follicular lymphoma (FL), and mantle cell lymphoma (MCL). Lymphoma is a form of cancer that begins in the lymphatic system, impacting lymphocytes, which are a specific type of white blood cell. This research addresses these challenges by proposing ensemble and non-ensemble transfer learning models employing pre-trained weights from VGG16, VGG19, DenseNet201, InceptionV3, and Xception. For the ensemble technique, this paper adopts a stack-based ensemble approach. It is a two-level classification approach and best suited for accuracy improvement. Testing on a multiclass dataset of CLL, FL, and MCL reveals exceptional diagnostic accuracy, with DenseNet201, InceptionV3, and Xception exceeding 90% accuracy. The proposed ensemble model, leveraging InceptionV3 and Xception, achieves an outstanding 99% accuracy over 300 epochs, surpassing previous prediction methods. This study demonstrates the feasibility and efficiency of the proposed approach, showcasing its potential in real-world medical applications for precise lymphoma diagnosis.