Packyanathan Ganesan
Research Subjects
Research Interests
Packyanathan Ganesan has a total of 97 co-authors. Here are the top co-authors by number of shared publications:
B. S. Sathish
25 shared publicationsG. Sajiv
Saveetha School of Engineering, Saveetha Institute of Medical and Technical Sciences, Saveetha University
15 shared publications- L. Maria Irudaya Leo Joseph
SR University, India
10 shared publications B.S. Sathish
Sri Venkateshwara College of Engineering,Dept. of Computer Science Engineering AI,Bengaluru,India
9 shared publicationsR. Murugesan
Narshima Reddy Engineering College
9 shared publicationsV. Rajini
Department of Electrical and Electronics, SSN College of Engineering, Chennai
8 shared publications
Latest Publications
- 31 August 2025
- Published by Springer Nature
- p. 415-427
Cataract Identification and Classification Using Machine Learning Algorithms
- 31 August 2025
- Published by Springer Nature
- p. 415-427
Cataract Identification and Classification Using Machine Learning Algorithms
- 31 August 2025
- Published by Springer Nature
- p. 401-413
An Approach to Salt and Pepper Noise Reduction in Color Images Using Deep Learning-Based Denoising Convolutional Neural Network
- 31 August 2025
- Published by Springer Nature
- p. 401-413
An Approach to Salt and Pepper Noise Reduction in Color Images Using Deep Learning-Based Denoising Convolutional Neural Network
- 4 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 156-162
An Innovative Framework for Early Detection and Classification of Diabetes: A Machine Learning Perspective
The global health crisis of diabetes demands precise diagnostic solutions which protect both individual and healthcare system outcomes. This research introduces an advanced model-based approach for diabetes diagnosis during early stages alongside classification methods to raise diagnostic precision and reliability. This research examines seven machine learning algorithms starting with Random Forest (RF) and AdaBoost (AB) alongside Decision Tree (DT) and kNearest Neighbor (kNN), followed by Naive Bayes (NB) along with Logistic Regression (LR) and Support Vector Machine (SVM). The performance measures accuracy alongside precision and recall together with F1-score of the system showed how the models performed. Experimental results confirm Random Forest's superiority as a good approach for diabetes classification and demonstrated its supremacy among all analytical metrics in this research. SVM showed the minimum performance due to its constraints when analyzing the dataset used. Evaluation outcomes demonstrate excellent resilience among ensemble learning methods including Random Forest and AdaBoost as opposed to standard classifiers such as Naive Bayes and Logistic Regression. Through this research we understand how machine learning revolutionizes diabetes care by examining the best approaches to building performance-enhanced diagnostic systems that benefit the healthcare industry.
- 4 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 156-162
An Innovative Framework for Early Detection and Classification of Diabetes: A Machine Learning Perspective
The global health crisis of diabetes demands precise diagnostic solutions which protect both individual and healthcare system outcomes. This research introduces an advanced model-based approach for diabetes diagnosis during early stages alongside classification methods to raise diagnostic precision and reliability. This research examines seven machine learning algorithms starting with Random Forest (RF) and AdaBoost (AB) alongside Decision Tree (DT) and kNearest Neighbor (kNN), followed by Naive Bayes (NB) along with Logistic Regression (LR) and Support Vector Machine (SVM). The performance measures accuracy alongside precision and recall together with F1-score of the system showed how the models performed. Experimental results confirm Random Forest's superiority as a good approach for diabetes classification and demonstrated its supremacy among all analytical metrics in this research. SVM showed the minimum performance due to its constraints when analyzing the dataset used. Evaluation outcomes demonstrate excellent resilience among ensemble learning methods including Random Forest and AdaBoost as opposed to standard classifiers such as Naive Bayes and Logistic Regression. Through this research we understand how machine learning revolutionizes diabetes care by examining the best approaches to building performance-enhanced diagnostic systems that benefit the healthcare industry.
- 2 April 2025
- Published by Springer Nature
- p. 245-257
Deep Learning Based Speckle Noise Removal in Color Images Using Supervised Denoising Convolutional Neural Network
- 2 April 2025
- Published by Springer Nature
- p. 245-257
Deep Learning Based Speckle Noise Removal in Color Images Using Supervised Denoising Convolutional Neural Network
- 21 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Comprehensive Machine Learning Approach for Wild Fire Detection and Classification Using Squeezenet and Ridge Regularized Logistic Regression
Wild fire is a serious environmental and socioeconomic menace and therefore any effort aimed at the early and accurate detection of wild fire has to be very useful. This paper proposes a novel machine learning framework for wildfire detection as well as its classification utilizing SqueezeNet as a lightweight convolutional neural network (CNN) and ridgeregularized logistic regression (RR-LR) as a classifier to improve the classification performance and model stability. The proposed method is compared with a set of machine learning classifiers like kNN. Al deliberation of these experimentation results suggests that the ridge-regularized logistic regression model performs better than these methods, specifically in cases concerning overfitting and in situations where the amount or balance of data available is small. Furthermore, SqueezeNet achieves efficient feature extraction that directly lessens the computation cost so users can deploy it in real-time especially in areas with limited resources. This work reveals the benefits of combining deep feature extraction with proper regularization methods and points to a promising solution for the wildfire detection and classification problem.
- 21 March 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
A Comprehensive Machine Learning Approach for Wild Fire Detection and Classification Using Squeezenet and Ridge Regularized Logistic Regression
Wild fire is a serious environmental and socioeconomic menace and therefore any effort aimed at the early and accurate detection of wild fire has to be very useful. This paper proposes a novel machine learning framework for wildfire detection as well as its classification utilizing SqueezeNet as a lightweight convolutional neural network (CNN) and ridgeregularized logistic regression (RR-LR) as a classifier to improve the classification performance and model stability. The proposed method is compared with a set of machine learning classifiers like kNN. Al deliberation of these experimentation results suggests that the ridge-regularized logistic regression model performs better than these methods, specifically in cases concerning overfitting and in situations where the amount or balance of data available is small. Furthermore, SqueezeNet achieves efficient feature extraction that directly lessens the computation cost so users can deploy it in real-time especially in areas with limited resources. This work reveals the benefits of combining deep feature extraction with proper regularization methods and points to a promising solution for the wildfire detection and classification problem.
- 4 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 380-385
An Efficient Chrominance Based Detection and Segmentation of Rotten Pyrus Malus Using a Cooperative Approach of Hill Climbing Optimization and Modified FCM Clustering
In this study, a novel method for the accurate and reliable identification and segmentation of rotten regions in Pyrus Malus (apples) is presented. In the proposed method, the initial cluster centroids computed from 3D color histogram of CIELab image. These centroids are utilized and processed using Hill Climbing (HC) Optimization to get the finest promising end result. Afterwards, the improved centroids are utilized by Modified FCM (MFCM) clustering method, which effectively divides the apple's rotten regions, if any. This collaborative strategy has a number of benefits: 1. The chrominance exploration increases overall efficacy by reducing susceptibility to changes in lighting. 2. In improving arrangement of FCM, HC optimization enhances optimal cluster centers; consequently, leads to enhanced identification of the rotten area. 3. In conjunction with each other, the combined method helps to make further identification of rotting areas easy and facilitate the creation of computerized classification and quality checks. The results of the experiments give evidence that the mentioned strategy allows for obtaining better precision in identifying and classification of rotten areas in apples in comparison to the approaches based on the use of RGB. This strategy may be the best tool for sorting and grading of apples using the automated grading systems for optimizing the processing of food products in the sector.
- 4 October 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 380-385
An Efficient Chrominance Based Detection and Segmentation of Rotten Pyrus Malus Using a Cooperative Approach of Hill Climbing Optimization and Modified FCM Clustering
In this study, a novel method for the accurate and reliable identification and segmentation of rotten regions in Pyrus Malus (apples) is presented. In the proposed method, the initial cluster centroids computed from 3D color histogram of CIELab image. These centroids are utilized and processed using Hill Climbing (HC) Optimization to get the finest promising end result. Afterwards, the improved centroids are utilized by Modified FCM (MFCM) clustering method, which effectively divides the apple's rotten regions, if any. This collaborative strategy has a number of benefits: 1. The chrominance exploration increases overall efficacy by reducing susceptibility to changes in lighting. 2. In improving arrangement of FCM, HC optimization enhances optimal cluster centers; consequently, leads to enhanced identification of the rotten area. 3. In conjunction with each other, the combined method helps to make further identification of rotting areas easy and facilitate the creation of computerized classification and quality checks. The results of the experiments give evidence that the mentioned strategy allows for obtaining better precision in identifying and classification of rotten areas in apples in comparison to the approaches based on the use of RGB. This strategy may be the best tool for sorting and grading of apples using the automated grading systems for optimizing the processing of food products in the sector.
- 26 July 2024
- Published by Springer Nature
- p. 465-474
An Efficient Approach to Skin Color Segmentation Using Perceptually Uniform CIELab Color Model
- 26 July 2024
- Published by Springer Nature
- p. 465-474
An Efficient Approach to Skin Color Segmentation Using Perceptually Uniform CIELab Color Model