Browsing by Author "Weerasinghe, L"

Now showing 1 - 16 of 16

Embargo
ARChem: Augmented Reality Chemistry Lab
(IEEE, 2021-12-06) Menikrama, M. R. L. Y; Liyanagunawardhana, C. S; Amarasekara, H. G. D. M. I; Ramasinghe, M. S; Weerasinghe, L; Weerasinghe, I
One of the technologies that has been gaining ground in recent years is Augmented Reality (AR), which allows to insert virtual objects into a real-world view using a device's camera and screen. This form of interaction associated with education can improve teaching and experiencing practical knowledge in schools, especially in more difficult subjects such as Chemistry. This study focused on virtual education by providing a platform for students to follow practical oriented subjects like Chemistry. As a result, a mobile application is created with four main functions that assist students during their learning process of Chemistry using the AR technique. The main functions are, AR with Artificial Intelligence (AI), Chemical equation identification and correction with Image Processing, Chabot with sentiment analysis and text summarization. The application is developed by using Machine Learning, AI with Deep Learning and Mobile Application development technologies. ARChem shows 3D models of flasks with important descriptions with the use and also features a Chabot with text summarization for frequently asked questions.
Embargo
ARChem: Augmented Reality Chemistry Lab
(IEEE, 2021-10-27) Menikrama, M. R. L. Y; Liyanagunawardhana, C. S; Amarasekara, H. G. D. M. I; Ramasinghe, M. S; Weerasinghe, L; Weerasinghe, I
One of the technologies that has been gaining ground in recent years is Augmented Reality (AR), which allows to insert virtual objects into a real-world view using a device's camera and screen. This form of interaction associated with education can improve teaching and experiencing practical knowledge in schools, especially in more difficult subjects such as Chemistry. This study focused on virtual education by providing a platform for students to follow practical oriented subjects like Chemistry. As a result, a mobile application is created with four main functions that assist students during their learning process of Chemistry using the AR technique. The main functions are, AR with Artificial Intelligence (AI), Chemical equation identification and correction with Image Processing, Chabot with sentiment analysis and text summarization. The application is developed by using Machine Learning, AI with Deep Learning and Mobile Application development technologies. ARChem shows 3D models of flasks with important descriptions with the use and also features a Chabot with text summarization for frequently asked questions.
Embargo
Arogya -An Intelligent Ayurvedic Herb Management Platform
(IEEE, 2020-11-04) Pathiranage, N; Nilfa, N; Nithmali, M; Kumari, N; Weerasinghe, L; Weerathunga, I
Ayurvedic means a science of life and well-being with its unique approaches to social and spiritual life. Especially in Sri Lanka we have our own set of rare Ayurvedic herbs which have been utilized by generations as medicinal treatments for a variety of diseases. Absence of specialists in this area makes proper identification as well as classification of valuable herbal plants a tedious task, which is essential for better treatment. Hence, a fully automated system for herb detection and classification, information visualization regarding them is highly desirable. There are existing applications which can identify plants with low prediction accuracies, as well as to give information regarding them. However, these applications are based on foreign plant data sets that do not include valuable herbs and shrubs with medicinal qualities. Hence this research proposes an application unique to medicinal plants, which can perform all these functionalities in both online and offline approach. Here, a new Ayurvedic plant dataset prepared from scratch, and preliminary results for classification of 5 types of herbs, compared with several deep Convolutional Neural Network (CNN) models based on transfer learning are presented. Experimental results indicate Marker-based Watershed algorithm as the best object detection algorithm in a complex background, VGG-16 as the best deep CNN classification model which reached a promising testing accuracy of 99.53%, and Seq2Seq LSTM model as the best deep learning model with optimum accuracy in abstractive information summarization.
Embargo
Arogya-An Intelligent Ayurvedic Herb Management Platform
(IEEE, 2020-10-15) Pathiranage, N; Nilfa, N; Nithmali, M; Kumari, N; Weerasinghe, L; Weerathunga, I
Ayurvedic means a science of life and well-being with its unique approaches to social and spiritual life. Especially in Sri Lanka we have our own set of rare Ayurvedic herbs which have been utilized by generations as medicinal treatments for a variety of diseases. Absence of specialists in this area makes proper identification as well as classification of valuable herbal plants a tedious task, which is essential for better treatment. Hence, a fully automated system for herb detection and classification, information visualization regarding them is highly desirable. There are existing applications which can identify plants with low prediction accuracies, as well as to give information regarding them. However, these applications are based on foreign plant data sets that do not include valuable herbs and shrubs with medicinal qualities. Hence this research proposes an application unique to medicinal plants, which can perform all these functionalities in both online and offline approach. Here, a new Ayurvedic plant dataset prepared from scratch, and preliminary results for classification of 5 types of herbs, compared with several deep Convolutional Neural Network (CNN) models based on transfer learning are presented. Experimental results indicate Marker-based Watershed algorithm as the best object detection algorithm in a complex background, VGG-16 as the best deep CNN classification model which reached a promising testing accuracy of 99.53%, and Seq2Seq LSTM model as the best deep learning model with optimum accuracy in abstractive information summarization.
Embargo
Betel Plant Tech: Betel Disease Forecasting System and Finding Marketplace
(IEEE, 2022-12-26) Maitipe, C.N; Bandara, G.P.C.C; Anuradi, M.R.C; Marambage, M. H. B. Y; Weerasinghe, L
Betel in Sri Lanka extends back to 340 B.C. and it has a significant cultural value in Sri Lanka. Betel is currently planted throughout the country, and it is the primary source of income for numerous farmers. Betel leaves are easily exposed to diseases. Taking that into consideration, it is clear that anticipating the spread of betel diseases, which have been identified as having a significant influence on the country’s economy, and creating a better marketplace is critical. The "Betel Plant Tech" smartphone application was created in this project with a focus on image processing, sequential models, regression and classification, in machine learning techniques. The findings indicate that predictions of betel yield identification produces high accuracy of 98% in random forest regression (RFR), identification of betel diseases with 92% accuracy with Restnet 34, and disease propagation level consists of a 92% accuracy level in the sequential model while predicting the spread of viral/fungal diseases has a 97% accuracy rate in Decision Tree Regression. The planned study would determine and predict the yield of betel cultivation, identify the diseases of betel leaf, predict the prevalence of diseases, and identify the level of disease prevalence. Besides, it will be easy to find a good marketplace for betel growers.
Embargo
CEYLAGRO: Information Technological Approach for an Optimized and Centralized Agriculiture Platform
(IEEE, 2020-12-10) Kaushalya, T. V. H; Wijewardana, B. Y. S; Karunasena, A; Kavishika, M. G. G; Gamage, S. T. A; Weerasinghe, L
Sri Lankan Agriculture sector can be considered as a crucial component as it contributes 18% of country GDP. As native farmers still cling to inapplicable traditional theorems and practices to track customer's vegetable consumption trends, they failed to assure a “good price” for their harvest. Also, the plants are prone to many diseases and pests' attacks which causes loss of the harvest. Unreliable problem identification, poor knowledge on application of fertilizers and pesticides have caused the farmers to lose their profits. As a solution to mitigate these problems, this study has built a computerized system with a vegetable price prediction system and a plant disease, pest identification system. Taking Potato as an example, the parameters of the time series model were analyzed through experiment and has built the price predictor using ARIMA model. Also, with advanced Image processing and CNN techniques Plant disease, pest identifier has built. Desirable results of the entire system have been achieved with more than 94%-97% rate of accuracy. The ultimate goal of this study is to achieve the optimal growth of the sector by navigating the users for a quality and effective decision making by reliable market trends and problem identification.
Embargo
Cocopal - A Deep Learning Based Intelligent System to Certify and Standardize the Quality of Coconut Based Products
(IEEE, 2022-12-09) Gunawardana, K.H.R.; Deshan, M.P.N.; Hemachandra, M.G.S.P.; Ganegoda, D; Hettiarachchi, N. M; Weerasinghe, L
The procedure of certifying and standardizing the quality of the coconut-based products is done manually in Sri Lanka at precent. It is a time consuming and labor-intensive task and is conducted by experts. In most cases, the quality is decided solely by visual inspections by buyers and suppliers, with no scientific basis. The paper reports the capacity of bringing modern technology solutions such as Artificial Intelligence (AI), Machine Learning (ML), Image Processing (IP), and decentralized storage to aid in the certification and standardization of the quality of raw materials.Results showed that the accuracy of the proposed system is in the 86% to 90% range and showed that this technique could beimproved and used as an alternative to manual techniques.
Embargo
CXR Scan:X-Ray Image Scanning Application for Lung Cancer and Tuberculosis
(Institute of Electrical and Electronics Engineers, 2022-10-15) Jayasooriya, A.M.U.J.; Wickramasekara, T.M.A.M; Jayasinghe, I.C.; Gunaratne, U.A.; Weerasinghe, L; Dassanayake, G. T
The initial criterion for identifying lung disorders is chest radiographs. The three major lung illnesses that pose the greatest threat to public health are tuberculosis, pneumonia, and lung cancer. Chest X-ray diagnosis of pulmonary illnesses is a challenging undertaking that requires high experience. In rural places, it can be difficult to locate skilled radiologists. Due to the high frequency of TB and lung cancer radiological similarities, many individuals with lung cancer are initially misdiagnosed as having TB and treated incorrectly. According to a recent WHO survey, millions of people die each year as a result of delayed or incorrect diagnoses of lung diseases. This death rate can be reduced, by early detection of certain disorders. This paper proposes a system with 4 main components; Image processing of chest X-rays to identify the disease using Convolutional Neural networks; Predicting the probability of having LC or TB using multivariate data classification techniques; Recommending medicine and related information to support the decision-making process using gaussian naïve bayes, logistic regression model and decision tree classification methods; Visualizing the X-ray image using Augmented Reality.
Embargo
Deep Learning Based Dog Behavioural Monitoring System
(IEEE, 2020-12-03) Boteju, W. J. M; Herath, H. M. K. S; Peiris, M. D. P; Wathsala, A. K. P. E; Samarasinghe, P; Weerasinghe, L
Dogs are one of the most popular pets in the world. It is usual that pet owners are always concerned about the health and the wellbeing of their pets. The activity levels of the dogs vary from each other based on breed and age. Tracking the behavioral changes using image processing and machine learning concepts and notifying the pet owners via a mobile application is the main objective of this research. Breed recognition has been done applying deep learning concepts to the user-uploaded video or the photograph of the dog. This research mainly focuses on walking, running, resting, and barking activity patterns of the dog. A surveillance camera and sensors were the main equipment for data collection. The audio feature of the surveillance camera is used to identity the barking behavior of the dog. Dogs from different ages belonging to Pomeranian and German Shepherd breeds have been selected for this experiment. Transfer learning with ResNet50, Inception V3, and support vector machines have been used to recognize and classify the activities of the dogs. The research study was able to achieve the accuracy levels as follows: - breed recognition - 89%+, walking pattern recognition - 99.5%, resting pattern recognition - 97% and barking pattern recognition - 60%. With the above accuracy levels, the research was able to identify the unusual behaviour of the dogs.
Open Access
An Efficient Automated Attendance Entering System by Eliminating Counterfeit Signatures using Kolmogorov Smirnov Test
(Global Journal, 2019-05-27) Weerasinghe, L; Sudantha, B. H
Maintaining the attendance database of thousands of students has become a tedious task in the universities in Sri Lanka. This paper comprises of 3 phases: signature extraction, signature recognition, and signature verification to automate the process. We applied necessary image processing techniques, and extracted useful features from each signature. Support Vector Machine (SVM), multiclass Support Vector Machine and Kolmogorov Smirnov test is used to signature classification, recognition, and verification respectively. The described method in this report represents an effective and accurate approach to automatic signature recognition and verification. It is capable of matching, classifying, and verifying the test signatures with the database of 83.33%, 100%, and 100% accuracy respectively
Embargo
Intelligent System for Skin Disease Detection of Dogs with Ontology Based Clinical Information Extraction
(Institute of Electrical and Electronics Engineers Inc., 2022-10-29) Rathnayaka, R. M. N. A; Anuththara, K. G. S. N; Wickramasinghe, R.J.P; Gimhana, P. S; Weerasinghe, L; Wimalaratne, G
The largest organ in dogs, the epidermis, is crucial in supplying immunological responses. Skin will preserve all the nutrients and safeguard the cells while warding off harmful or pathogenic substances. Most dog owners today are not aware that their pet dog has a skin condition. Although they were aware of these ailments, they had no notion of how to cure them. In such a situation, the dog may experience pain and an aggravation of the condition. Owners should therefore take their dogs to the vet, even if the skin condition is minor. It can, however, be a costly procedure. There aren't many forums where dog owners may get advice from professionals and ask inquiries regarding their pets. The solution suggests a fully functional mobile application which is a combination of disease identification feature, disease severity level detection feature, domain specific knowledge base with semantic web development and a domain specific AI based chat-bot to the dog owners to overcome this problem using Convolutional Neural Network (CNN) and natural language processing (NLP).System will extract the necessary features from the images of the lesion to classify the skin condition and Severity level of the disease. The results obtained show disease type classification is within the accuracy range of 77.78% to 100% which tested again 4 CNN base models. As for the severity level identification accuracy situated around 99.62%.
Embargo
An Interactive E-Learning Tool
(IEEE, 2022-07-18) Kodagoda, D. G; Ishara, K.G.R.U; Kumara, R. M. R. P; Dilshan, W. A. D.T; Weerasinghe, L; Premadasa, N
Use of E-learning systems has surged immensely during the Covid-19 pandemic, which started around 2020. This research specifically conducted to introduce novel features with the purpose of enhancing traditional E Learning platforms. The suggested features are namely, avoid unauthorized users from accessing private video sessions using face recognition, manipulating 3D objects by hand gestures, analyzing student’s attention using face landmarks, smart QAs using voice recognitions. These features will provide not only an enhancement for e-Learning platforms but also it will improve user experience, efficiency, and effectiveness of current tools up to a certain distinguishable level.
Embargo
OMNISCIENT: A Branch Monitoring System for Large-scale Organizations
(IEEE, 2020-12-10) Jayasekara, T; Omalka, K; Hewawelengoda, P; Kanishka, C; Samarasinghe, P; Weerasinghe, L
Omniscient is a system that enables higher-level management of massive organizations to remotely monitor and scrutinize the activities that take place in the branches from the head office itself by providing exclusive insight in the form of detailed reports on the employees' behaviour and performance daily, weekly and monthly. The system further monitors the branch and provides reports on any suspicious behaviour and also on the customers' activity within the branch premises. Omniscient rates the customer's level of satisfaction by capturing the customer's facial expressions and analyzing their emotions while they are being served. The employee face and dress recognition models have accuracies of 90.90% and 87.00% respectively while, employee activity detection has an accuracy of 89.00%. Customer emotion and miscellaneous activities detection models have the accuracies of 91.50% and 83.00% respectively. All of the aforementioned procedures were made possible by systematically analyzing the IP camera video footage obtained throughout the day to analyze the work productivity and performance of the branch as accurately as possible using deep learning and modern visual computing techniques like CNN, OpenCV, Haar Cascade classifier, face recognition, Dlib and Darknet.
Embargo
Plagiarism Detection Tool for Enhanced Entity-Relationship Diagrams
(IEEE, 2021-12-01) Dahanayake, H; Samarajeewa, D; Jayathilake, A; Bandara, D; Karunasena, A; Weerasinghe, L
Plagiarism is presenting someone else’s work as one’s own work without giving credit to the original owner. Recently, plagiarism has become a serious issue in the fields of Education and Technology. To address this issue, many systems have been implemented to detect plagiarism. However, most of them are designed to deal with plagiarism of text content. Detecting plagiarism in figures and diagrams is equally important. Although there is research done on detecting plagiarism in images and flow charts, there is no research done on detecting plagiarism in more complex diagrams such as Enhanced Entity-Relationship (EER) diagrams. This paper presents a methodology to detect plagiarism in EER diagrams using Deep Neural Networks (DNN), image processing techniques, Optical Character Recognition (OCR) techniques, and text similarity detection algorithms. Since the students are aware of the existence of a plagiarism detecting tool, it will encourage the students to do work on their own and it will reduce exam offenses. The similarity report can be presented as proof to the offenders who are not accepting that they have plagiarized others' work. Using the proposed system, the EER diagram plagiarism can be detected much faster and accurately. Therefore, the efficiency of marking examinations will be increased. The final outcome of the system will be a similarity report including the plagiarized content in the compared EER diagrams.
Embargo
SMARKET-Shopping in Supercenters (Hypermarkets) with Augmented Reality
(IEEE, 2021-12-17) Jayagoda, N. M; Jayawardana, O. R; Welivita, W. W. T. P; Weerasinghe, L; Dassanayake, T
Not so long ago, online shopping for groceries, electronics, and furniture items seemed futuristic. But today, it has become a norm to order requisites through online platforms using smart devices and deliver them to customers' doorstep. With the emerge of technologies such as artificial intelligence, machine learning, deep learning, augmented reality, retail becomes progressively effortless. One such emerging futuristic technology involved recently in online shopping is Augmented Reality (AR) which is rapidly adopted by many industries. In multi-story supercenters, also known as “Hypermarkets”, the customer often feels lost due to difficulty in finding exactly what they looking for, and also in conventional online shopping, often customers are in two minds whether to purchase an item or not since it lacks the proper visualization, touch, and feel of the product. In this research study, we propose a mobile-based solution with augmented reality, which assists the customer when shopping in-store as well as when shopping online to mitigate the difficulties and hesitancies faced while shopping. The results are commendable with 96.21 % accuracy in suggesting visually similar items and 89.59% accuracy in detecting emotional implications of product reviews.
Embargo
Yuwathi: Early Detection of Breast Cancer and Classification of Mammography Images Using Machine Learning
(IEEE, 2022-07-18) Diddugoda, D; Fernando, D. B; Munasinghe, S. M; Weerasinghe, L; Weerathunga, I
According to the World Health Organization's (WHO) data and records, breast cancer is one of the most common diseases among women. As a result of the mutations of the genes within a cell, the cell starts growing uncontrollably and rapidly. Such a condition is known as cancer. Cancer tumors can be categorized into two major categories, benign and malignant. However, there is no existing solution in practice to automate early breast cancer identification and risk prediction using medical images (Mammograms). This paper discusses automating breast cancer detection, breast density identification, risk prediction, and solution suggestion using machine learning, image processing, and computer vision techniques. All the mentioned features can be accessed using the application "YUWATHI", and a user can take advantage of this application by using a smartphone also a web application. The objectives of the present study are mammographic mass detection without user intervention, identifying pectoral muscles and removing them, training a machine learning model to identify the future risk of breast cancers by obtaining clinical reports from the OCR application and suggesting solutions for the above problems using a computer-aided diagnosis (CADx) system that helps doctors to make decisions swiftly. The algorithms used for breast cancer detection, breast density classification, and future breast cancer risk prediction are Convolutional Neural Network (CNN), CNN and Logistic Regression with the accuracies 97.32%, 71.97% 74.76%, respectively.