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

Filters

2022 - 20266

Advanced Search

Filter by

Settings

Start date: 2022Subject: search.filters.subject.computer vision

Search Results

Now showing 1 - 6 of 6
  • Thumbnail Image
    ItemEmbargo
    AgriSense LK: Weekly Automated Machine Learning for Sri Lankan Produce Prices with Business Continuity Plan, Market Opportunity Ranking, Cultivation Targeting, and Yield Quality Valuation
    (Institute of Electrical and Electronics Engineers Inc., 2026-05-22) Matharaarachchi, Charaka J.; Samarasinghe, Ravindu T; Vidyasarani G.G.T.; Fasnas, M; Siriwardana, D; Wijesooriya, A
    In Sri Lanka, agricultural decision-making remains largely traditional: decisions are often based on historical practices, informal consultation, and heuristic judgment. The primary barrier is that market price data is difficult to interpret without analytical expertise, resulting in unpredictable price volatility and suboptimal farmer income. AgriSense LK is a machine learning platform that converts historical price records into actionable recommendations for farmers and traders. The system comprises four components: business strategy classification, market opportunity ranking, cultivation targeting, and smartphone-based produce quality grading. The platform was trained on 123,985 real price records sourced from the Central Bank of Sri Lanka (CBSL), spanning 2017 to 2025. Key results include a MAPE of 0.7% and MAE of Rs. 1.86 on weekly price forecasting (a 98.1% improvement over the naive baseline), a ROC-AUC of 0.9056 on cultivation targeting, and 91.49% crop classification accuracy with 89.84% quality grade accuracy in the computer vision component. Direct price regression over a seven-day horizon proved unreliable; a binary profitability classifier was adopted instead and substantially outperformed the regression approach. While results are promising, further validation under real-world deployment conditions is required.
  • Thumbnail Image
    ItemEmbargo
    Automated Bilingual Handwriting Evaluation for Early Childhood Education: A Multi-Metric Structural Analysis
    (Institute of Electrical and Electronics Engineers Inc, 2026-07-02) Gunawardena, T; Rathnasinghe, E; Abesinghe, C; Mihara, B; Abeywardhana, L; Weerasinghe, L; Selvaratnam, N
    Traditional handwriting assessment in early childhood education relies on subjective teacher evaluation and becomes particularly challenging in bilingual Sri Lankan curricula involving both English and Sinhala. This paper introduces a multi-metric handwriting quality scoring pipeline for automated evaluation of preschool handwritten letters for children aged 4-6. The proposed system integrates contour-based segmentation, language-specific convolutional neural network (CNN)-based recognition (achieving 86.73% test accuracy on EMNIST for English and 99.04% on a custom Sinhala dataset), and an eight-metric structural analysis framework. The framework combines structural similarity (SSIM), topology preservation, shape invariants, proportion analysis, stroke continuity, alignment, pixel Intersection over Union (IoU), and a 9 × 9 grid-based spatial comparison. The weighted scoring approach shows strong agreement with teacher assessments, demonstrating the reliability of the proposed system. Designed for Sri Lankan early education contexts, the approach provides a scalable, interpretable, and objective solution, addressing the gap in automated Sinhala handwriting assessment tools.
  • Thumbnail Image
    PublicationOpen Access
    Advancing Object Detection: A Narrative Review of Evolving Techniques and Their Navigation Applications
    (Institute of Electrical and Electronics Engineers Inc., 2025-03-17) Tennekoon, S; Wedasingha, N; Welhenge, A; Abhayasinghe, N; Murray Am, I
    Object detection plays a pivotal role in advancing computer vision systems by enabling machines to perceive and interact intelligently with their environments. Despite significant advancements, comprehensive exploration of its evolution and applications in navigation remains underrepresented. This review paper examines the evolution of object detection technologies, from early methodologies to contemporary advancements, and their critical role in navigation tasks. The emphasis was on the significance of contextual learning in enhancing object detection performance by leveraging spatial and temporal information. Furthermore, the limitations of conventional approaches that rely heavily on hand-engineered features are examined. It is then demonstrated that contextual learning facilitates automated feature extraction, resulting in improved accuracy exceeding a 50% increase and adaptability in diverse applications. The review concludes by outlining future trends and opportunities for further advancements in object detection and, underscoring its transformative impact on autonomous navigation and beyond. In summary, this review contributes to a comprehensive understanding of object detection technologies by offering insights into their evolution, highlighting their applications in navigation, and providing guidance for future research in context-aware systems.
  • Thumbnail Image
    PublicationOpen Access
    POSTUREEASE: A Web Based Application for Monitoring the Sitting Posture in Computer Based Working Environment
    (SLIIT City UNI, 2025-07-08) Thennakoon, T.M.C.L; Worthington, A.E
    In today’s digital era, prolonged computer usage is commonplace, particularly in professional environments. However, extended periods of improper sitting posture can result in musculoskeletal disorders, fatigue, and chronic health complications. Addressing this concern, this research presents PostureEase, a web-based posture analysis application designed to promote ergonomic awareness and encourage healthy sitting habits. The system leverages computer vision and machine learning technologies to monitor posture in real time using webcam input. Developed with a React-based frontend and a Python-Flask backend, PostureEase processes live video streams through OpenCV and MediaPipe to detect poor posture based on facial and shoulder landmarks. Upon detecting improper alignment, the system provides immediate alerts to the user. Key features include posture history tracking, automated report generation, and exercise and ergonomic recommendations. Evaluation of the system demonstrated reliable performance under typical working conditions, with responsive detection and user-friendly interaction. This research contributes to the domain of health technology by offering a practical and preventive tool for posture correction. Future enhancements may include mobile integration and personalized analytics to further improve user experience and effectiveness. With a modular architecture and high usability, PostureEase achieved an accuracy of 92% in posture classification under normal lighting and device conditions. The system was evaluated through both user testing and technical validation, highlighting its potential for scalable deployment in ergonomic health monitoring.
  • Thumbnail Image
    PublicationEmbargo
    Automated vehicle insurance claims processing using computer vision, natural language processing
    (IEEE, 2022-11-30) Fernando, N; Kumarage, A; Thiyaganathan, V; Hillary, R; Abeywardhana, L
    Traditional insurance claims processing systems are no match for the modern world due to the increasing population of vehicles and the resulting number of accidents. In this paper, the authors present a novel idea to automate the tedious processes in the insurance industry. The presented system consists of three main components namely, re-identify the make and model of the vehicle, identify the damaged automobile component, type, and severity, and compute an accurate repair estimate using damage component identification. Also, automate the documentation process by identifying the relevant fields in the voice input provided by the user. This ensures both the parties involved in this process will be benefited from the proposed system. Presented solutions Were designed using the aid of Artificial Intelligence techniques, mainly CNN models and Natural language processing techniques.
  • Thumbnail Image
    PublicationOpen Access
    Real-Time Embedded System for Inattentive Driver Monitoring
    (SLIIT, 2022-02-11) Nalmi, R; Clerence, A.; Buddhika, P; Saranyan
    One of the causes of motor vehicle accidents in Sri Lanka is driver inattention or drowsiness. In the field of intelligent transportation systems, continuous research and development are conducted to address this contemporary issue. Many approaches, such as driver assistance and drowsiness detection systems, have been proposed to overcome this fatality. The purpose of this research was to implement a product that can maximise road safety while improving the transport sector's efficiency and reliability of the logistics chain to reinforce the country's economic growth. In this paper, the correlation between the preprocessed vehicular parameters and visual features are used to analyse the driver state and make predictions of the driver's perfomance. The proposed system uses computer vision and fuzzy logic inference implemented on the singleboard computer Raspberry Pi to detect facial features and to determine the driver's drowsiness state, an ELM327 is used to read the vehicle parameters from the Electronic Control Unit (ECU) and motion sensors were used to obtain the steering angle. The data acquired is stored in a cloud platform using REST API. The database also contains driver details. The system uses a fingerprint scanner to identify the driver. An actuator was installed in the vehicle to alert the driver when the system detects inattentiveness. Overall the proposed project provided satisfying experimental results. It can be used as a solution to improve road safety and a supporting tool for the logistics sector to monitor vehicles and driver performance.