Research Papers - Department of Electrical and Electronic Engineering

Permanent URI for this collectionhttps://rda.sliit.lk/handle/123456789/679

Browse

Filters

2021 - 20254

Advanced Search

Filter by

Settings

Author: search.filters.author.Rathnayake, NStart date: 2020

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    PublicationOpen Access
    Efficient Hotspot Detection in Solar Panels via Computer Vision and Machine Learning
    (Multidisciplinary Digital Publishing Institute (MDPI), 2025-07-15) Fernando, N; Seneviratne, L; Weerasinghe, N; Rathnayake, N; Hoshino, Y
    Solar power generation is rapidly emerging within renewable energy due to its cost-effectiveness and ease of deployment. However, improper inspection and maintenance lead to significant damage from unnoticed solar hotspots. Even with inspections, factors like shadows, dust, and shading cause localized heat, mimicking hotspot behavior. This study emphasizes interpretability and efficiency, identifying key predictive features through feature-level and What-if Analysis. It evaluates model training and inference times to assess effectiveness in resource-limited environments, aiming to balance accuracy, generalization, and efficiency. Using Unmanned Aerial Vehicle (UAV)-acquired thermal images from five datasets, the study compares five Machine Learning (ML) models and five Deep Learning (DL) models. Explainable AI (XAI) techniques guide the analysis, with a particular focus on MPEG (Moving Picture Experts Group)-7 features for hotspot discrimination, supported by statistical validation. Medium Gaussian SVM achieved the best trade-off, with 99.3% accuracy and 18 s inference time. Feature analysis revealed blue chrominance as a strong early indicator of hotspot detection. Statistical validation across datasets confirmed the discriminative strength of MPEG-7 features. This study revisits the assumption that DL models are inherently superior, presenting an interpretable alternative for hotspot detection; highlighting the potential impact of domain mismatch. Model-level insight shows that both absolute and relative temperature variations are important in solar panel inspections. The relative decrease in “blueness” provides a crucial early indication of faults, especially in low-contrast thermal images where distinguishing normal warm areas from actual hotspot is difficult. Feature-level insight highlights how subtle changes in color composition, particularly reductions in blue components, serve as early indicators of developing anomalies.
  • Thumbnail Image
    PublicationEmbargo
    Performance Comparison of Sea Fish Species Classification using Hybrid and Supervised Machine Learning Algorithms
    (IEEE, 2022-10-04) Nalmi, R; Rathnayake, N; Mampitiya, L.I
    In the domain of autonomous underwater vehicles, the classification of objects underwater is critical. The hazy effect of the medium causes this obstacle, and these effects are directed by the dissolved particles that lead to the reflecting and scattering of light during the formation process of the image. This paper mainly focuses on exploring the best possible image classifier for the underwater images of the different fish species. The classifications were carried out by different hybrid and supervised machine learning algorithms such as Support Vector Machine (SVM), Random Forest (RF), Neural Networks (NN), Logistic Regression (LR), Decision Tree (DT), and Naive Bayes (NB). This study compares the algorithms’ accuracy and time and analyzes crucial features to decide the most optimal algorithm. Furthermore, the results of this paper depict that using dimension reduction methods such as PCA and LDA increases the accuracy of some algorithms. Random Forest was able to outperforms with a higher accuracy of 99.89% with the proposed feature extraction methods.
  • Thumbnail Image
    PublicationEmbargo
    An Efficient Ocular Disease Recognition System Implementation using GLCM and LBP based Multilayer Perception Algorithm
    (IEEE, 2022-08-03) Rathnayake, N; Mampitiya, L. I
    This research study is focused on the classification of ocular diseases by referring to a well-known dataset. The data is divided into seven classes: diabetes, glaucoma, cataract, normal, hypertension, age-related macular degeneration, pathological myopia, and other diseases/abnormalities. A Neural Network is used for the classification of diseases. In addition, the GLCM and LBP feature extracting methods have been used to carry out the feature extraction for the fundus images. This study compares five different ocular disease recognizing techniques. Moreover, the proposed model was evaluated regarding precision, recall, and accuracy. The proposed solution outperformed existing state-of-the-art algorithms, achieving 99.58% accuracy.
  • Thumbnail Image
    PublicationEmbargo
    Classification of Human Emotions using Ensemble Classifier by Analysing EEG Signals
    (IEEE, 2021-04-13) Mampitiya, L. I; Nalmi, R; Rathnayake, N
    This study is based on EEG brain wave classification of a well-known dataset called the EEG Brainwave Dataset. The dataset combines three classes such as positive, negative, and neutral. The classification is performed using an ensemble classifier that combines RF, KNN, DT, SVM, NB, and LR. The meta classifier is LR, while the other five algorithms work as the base classifiers. Furthermore, PCA is used as the dimension reduction method to increase the accuracy of the final output. The results are evaluated under 11 different parameters. Moreover, the accuracy of this study is compared with the seven other EEG emotion classification methods. The proposing method attained 99.25% of accuracy, outperforming the other state-of-the-art algorithms.