Research Publications
Permanent URI for this communityhttps://rda.sliit.lk/handle/123456789/4194
This main community comprises five sub-communities, each representing the academic contribution made by SLIIT-affiliated personnel.
Browse
50 results
Filters
Advanced Search
Filter by
Settings
Search Results
Publication Open Access Solar Hotspot Detection Using VHDL-Simulated Fixed-Point SVM: A Methodology Toward FPGA Realization(Faculty of Engineering, 2026-03) Fernando, N; Seneviratne, L; Weerasinghe, N; Rathnayake, N; Hoshino, YEarly detection of thermal hotspots in photovoltaic modules is critical to ensuring their efficiency, safety, and longevity. This study presents a complete end-to-end methodology for implementing a fixedpoint Medium Gaussian Support Vector Machine classifier using VHDL for a Field Programmable Logic Array. The approach begins with feature extraction from thermal images of healthy and defective solar panels, which focuses on MPEG-7 descriptors. The study shows that high impact for hotspot detection comes from blue chrominance contrast. A medium Gaussian SVM model is trained in MATLAB and converted to a fixed-point Q1.15 format for hardware compatibility. Key parameters, including support vectors, Lagrange multipliers, bias, and kernel scale, are extracted and verified in a custom Python environment to ensure numerical alignment with MATLAB results. The validated model is then implemented in synthesizable VHDL. It is verified using GHDL and the GNU Tool Kit waveform viewer, confirming bit-accurate hardware behaviour. Results show classification accuracy exceeding 99.3% with negligible performance loss due to quantization. The design achieves deterministic latency through an FSM-based structure and parallel feature processing for a 300-support vector and 222-feature system. This method enables low-power, real-time inference on a UAV-based edge platform, primarily focusing on drones.Publication Open Access Predicting Cognitive Test Performance from New Onset Behavioral and Personality Changes in Adults over 50 using Post-Selection Boosted Random Forest Classifier(Faculty of Engineering, 2025-09-09) Mervyn M.; Welhenge A.; Creese B.Mild Behavioral Impairment (MBI) refers to neuropsychiatric symptoms of various severity levels that might not be discovered by conventional psychiatric nosology. These symptoms should persist for more than six (06) months. MBI is typically observed in adults of age 50 and above. This study investigates the prediction of cognitive test performance of cognitive and behavioral changes in adults over 50 years of age using a post-selection boosted Random Forest (RF) Classifier. The baseline cognitive aging data of the Simple Reaction Time (SRT) metric and Mild Behavioral Impairment Checklist (MBI-C) from the ongoing PROTECT study in the United Kingdom was used to classify the participants’ cognitive ability into five classes. Using the post-selected boosted RF classifier, the study obtained an accuracy of 96.26% which was an improvement compared to the 95.52% accuracy obtained by the RF classifier. These findings suggest that machine learning-based prediction models can provide valuable insights into analyzing the cognitive decline of adults of a late age.Publication Open Access Towards Safer Elderly Care: A Convolutional Neural Network Solution for Fall Detection(Faculty of Engineering, 2025-09-09) Kalupahana R.W; Maduranga M.W.PAs modern life becomes increasingly busy, computer vision-based monitoring systems have become essential, particularly in elderly care. This paper presents the development of a robust fall detection system using deep learning techniques, specifically a convolutional neural network (CNN) that processes RGB images to accurately distinguish between fall and non-fall events. The model is trained and validated on a dataset categorized into two classes: fall and non-fall. By utilizing convolutional and pooling layers, CNN effectively learns hierarchical representations of the input data, capturing both low-level and high-level features crucial for accurate fall detection. The key stages of this approach include data acquisition, pre-processing, and model training. The model's performance is evaluated using precision, recall, and F1-score metrics, demonstrating high accuracy, which is further enhanced through data augmentation, pre-processing, and crossvalidation techniques. A confusion matrix analysis confirms the model's effectiveness in correctly classifying instances across both classes. The system also extends its capabilities to video analysis by extracting frames at 30-second intervals, ensuring continuous and comprehensive monitoring. This research highlights the potential of deep learning to enhance safety and care for the elderly, offering a reliable solution for real-time fall detection. The findings underscore the importance of integrating advanced technologies into healthcare, paving the way for future innovations in monitoring and assistance systems.Item Embargo AI-Driven Vehicle Valuation and Market Trend Analysis for Sri Lanka's Automotive Sector(Institute of Electrical and Electronics Engineers Inc., 2025) De Silva K.P.N.T.; Shehan H.A.; Jayawardhane A.S; Premarathne A.P.S.; Krishara, J; Wijendra, D.RThe automotive sector in Sri Lanka faces challenges in vehicle valuation accuracy and market trend analysis due to fluctuating prices, varying vehicle conditions, and environmental concerns. This paper presents an AI-driven vehicle valuation system integrating machine learning models for automated vehicle identification, damage detection, market trend analysis, and environmental sustainability assessments. Using deep learning techniques such as Convolutional Neural Networks (CNNs) and time-series models like Long Short-Term Memory (LSTM), the system delivers accurate valuation and market trend insights. Experimental results demonstrate 9 2% accuracy in damage classification and a mean absolute error (MAE) of 5.3% in repair cost estimation, supporting informed decision-making. This research bridges gaps in valuation transparency and sustainability in emerging automotive markets.Item Embargo Intelligent Detection of Corporate Targeted Phishing Emails - A Hybrid Approach Combining Deep Learning Models with Domain Anomaly Detection(Institute of Electrical and Electronics Engineers Inc., 2025) Seethawaka, R; Chathurya N.E.G; Chandrasiri D.K.W.G.G.T; Kavithma K.A.S; Fernando, H; Wijesooriya, AThis paper introduces a system designed to detect corporate-targeted phishing emails by combining two key strategies: advanced email content analysis and domain similarity analysis. The system first examines the text of emails using a hybrid deep learning model that merges modern language understanding techniques with sequential pattern recognition, achieving high accuracy in identifying phishing intent. Two models were tested - a standalone Bi-LSTM sequential model and a hybrid version(BERT - Bi-LSTM) with the hybrid model proving superior, scoring an F1 score of 0.97 compared to 0.93 for the standalone model. Second, the system verifies sender domains to detect spoofing attempts, such as subtle typos, homograph attacks or TLD/subdomain spoofing. This domain check reduces reliance on text analysis alone, helping analysts prioritize threats more effectively. Tested against a mix of legitimate and malicious domains, the domain module achieved near-perfect accuracy, minimizing false alarms. By integrating these approaches, the system addresses a critical gap in existing methods, which often focus on only one aspect of phishing (e.g., email content or URL features). This dual strategy ensures a more comprehensive defense, particularly against sophisticated attacks that use convincing language paired with fake domains. The combined model not only improves detection accuracy but also supports security teams by providing clear, actionable insights, making it practical for real-world corporate environments.Item Embargo A Dual-Branch CNN and Metadata Analysis Approach for Robust Image Tampering Detection(Institute of Electrical and Electronics Engineers Inc., 2025) Zakey, A; Bawantha, D; Shehara, D; Hasara, N; Abeywardena, K.Y; Fernando, HImage tampering has become a widespread issue due to the availability of advanced tools such as Photoshop, GIMP, and AI-powered technologies like Generative Adversarial Networks (GANs). These advancements have made it easier to create deceptive images, undermining their reliability and fueling misinformation. To address this growing problem, we propose a hybrid approach for image forgery detection, combining deep learning with traditional forensic techniques. Our study integrates a dual-branch Convolutional Neural Network (CNN) with handcrafted features derived from Error Level Analysis (ELA), noise residuals from the Spatial Rich Model, and metadata analysis to enhance detection capabilities. Metadata analysis plays a crucial role in identifying inconsistencies in image properties such as timestamps, geotags, and camera details, which often accompany tampered images. The CASIA dataset, a publicly available benchmark for tampered images, was used to train and evaluate the proposed model. After 30 epochs of training, the hybrid method achieved an accuracy of 95%, demonstrating its effectiveness in distinguishing between authentic and tampered images. This research highlights the advantages of combining deep learning models with traditional feature extraction methods and metadata analysis, offering a robust solution for detecting manipulated images. Our findings contribute to advancing image forensics by improving detection accuracy, even in cases involving sophisticated tampering methods driven by AI.Publication Open Access Smart Sorting and Grading Fruits based on Image Processing Techniques(SLIIT City UNI, 2025-07-08) Ahamed, A J S; Benorith, LThis paper presents the design and implementation of an automated apple sorting system that integrates machine vision techniques with embedded control for real-time classification and sorting of apples. The system employs a Raspberry Pi 4 as the primary processing unit, using a YOLOv11 model for fruit detection and classification, while an Arduino Nano manages weight measurement via a load cell. Real-time images of apples on a conveyor belt are captured, processed, and classified into four categories: Good Red, Good Green, Bad Red, and Bad Green. Sorting mechanisms, including servos and actuate based on classification results, with an integrated LCD and cloudbased Google Sheets providing monitoring and logging. The system demonstrates high classification accuracy and reliable sorting performance, offering a cost-effective solution for small to mid-scale agricultural applicationsPublication Open Access Eco-friendly mix design of slag-ash-based geopolymer concrete using explainable deep learning(Elsevier, 2024-09) Ranasinghe, R.S.S.; Kulasooriya, W.K.V.J.B; Perera, U.S; Ekanayake, I.U.; Meddage, D.P.P.; Mohotti, D; Rathanayake, UGeopolymer concrete is a sustainable and eco-friendly substitute for traditional OPC (Ordinary Portland Cement) based concrete, as it reduces greenhouse gas emissions. With various supplementary cementitious materials, the compressive strength of geopolymer concrete should be accurately predicted. Recent studies have applied deep learning techniques to predict the compressive strength of geopolymer concrete yet its hidden decision-making criteria diminish the end-users’ trust in predictions. To bridge this gap, the authors first developed three deep learning models: an artificial neural network (ANN), a deep neural network (DNN), and a 1D convolution neural network (CNN) to predict the compressive strength of slag ash-based geopolymer concrete. The performance indices for accuracy revealed that the DNN model outperforms the other two models. Subsequently, Shapley additive explanations (SHAP) were used to explain the best-performed deep learning model, DNN, and its compressive strength predictions. SHAP exhibited how the importance of each feature and its relationship contributes to the compressive strength prediction of the DNN model. Finally, the authors developed a novel DNN-based open-source software interface to predict the mix design proportions for a given target compressive strength (using inverse modeling technique) for slag ash-based geopolymer concrete. Additionally, the software calculates the Global Warming Potential (kg CO2 equivalent) for each mix design to select the mix designs with low greenhouse emissions.Publication Embargo IoT-Enabled Smart Solution for Rice Disease Detection, Yield Prediction, and Remediation(IEEE, 2023-06-26) Wanninayake, K.M.I.S; Bambaranda, L.G.S. W; Wickramaarachchi, T.I; Pathirana, U.C.S.L; Vidhanaarachchi, S; Nanayakkara, A.A.E.; Gunapala, K.R.D.; Sarathchandra, S.R.; Gamage, A.I; De Silva, D.ISri Lanka's rice cultivation is a vital industry supporting over 1.8 million cultivators and providing staple sustenance for 21.8 million people. According to Sri Lanka's Central Bank, rice cultivation contributed 2.7% to the country's GDP in 2020 [3]. Pests and diseases, particularly rice thrips damage and rice blast disease, are a challenge for the industry, as they cause yield loss. This paper describes an intelligent solution that aids stakeholders by detecting and classifying the disease, forecasting its dispersion, and providing remedies. The proposed solution is approached with deep learning techniques for real-time detection and classification of the disease, location tracking of infected areas, and pesticide application on the target. In addition, it predicts the spread of disease based on the locations of infected individuals. In addition, the solution enables Machine-learning algorithms to recommend appropriate rice varieties and predict yields. In controlled experiments utilizing data from Sri Lankan paddy fields, the proposed method obtained high accuracy rates of 89%-98% in identifying disease and rice varieties and yield prediction. This system has the potential to increase rice production and productivity, decrease yield loss, and benefit the Sri Lankan rice industry and producers.Publication Open Access COVID-19 symptom identification using Deep Learning and hardware emulated systems(Elsevier, 2023-06-28) Liyanarachchi, R; Wijekoon, J; Premathilaka, M; Vidhanaarachchi, SThe COVID-19 pandemic disrupted regular global activities in every possible way. This pandemic, caused by the transmission of the infectious Coronavirus, is characterized by main symptoms such as fever, fatigue, cough, and loss of smell. A current key focus of the scientific community is to develop automated methods that can effectively identify COVID-19 patients and are also adaptable for foreseen future virus outbreaks. To classify COVID-19 suspects, it is required to use contactless automatic measurements of more than one symptom. This study explores the effectiveness of using Deep Learning combined with a hardware-emulated system to identify COVID-19 patients in Sri Lanka based on two main symptoms: cough and shortness of breath. To achieve this, a Convolutional Neural Network (CNN) based on Transfer Learning was employed to analyze and compare the features of a COVID-19 cough with other types of coughs. Real-time video footage was captured using a FLIR C2 thermal camera and a web camera and subsequently processed using OpenCV image processing algorithms. The objective was to detect the nasal cavities in the video frames and measure the breath cycles per minute, thereby identifying instances of shortness of breath. The proposed method was first tested on crowd-sourced datasets (Coswara, Coughvid, ESC-50, and a dataset from Kaggle) obtained online. It was then applied and verified using a dataset obtained from local hospitals in Sri Lanka. The accuracy of the developed methodologies in diagnosing cough resemblance and recognizing shortness of breath was found to be 94% and 95%, respectively.