https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2407 2026-03-26T17:28:49Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/3155 Title: Species Speci c Primer Design for Unique Identi cation of Closely Related Species Authors: Tharanga, W.D.R. Abstract: Unique identi cation of closely related species is important when it comes to biomedi- cal and molecular biology research. For example Dengue virus has four serotypes(DEN-1, DEN-2, DEN-3 and DEN-4) and identifying the speci c serotype is important for its med- ical treatment. Conventional identi cation methods based on physiochemical properties could not uniquely distinguish each subtype due to the high degree of similarities in their interactions. DNA analysis can be employed for the purpose of unique identi cation, because of its capability to identify genetic information which makes one species di erent from another. Polymerase Chain Reaction(PCR) is a technique which has been proven to be e cient in DNA analysis of closely-related species. The primary requirement of PCR is a pair of special DNA fragments called "primers" which are used to mark the DNA fragment of interest. However, to use PCR for this purpose, we need to design primers which mark unique DNA fragment(s) of a species which can distinguish it from the rest of the closely- related species. This research focuses on nding a computational method to generate species-speci c unique primers. PCR experiments could possibly fail due to several reasons such as self-dimer and hair- pin formation. Recent research from other classes of primer design (eg. general primer design, conserved primer design) have addressed this issue by using computational primer validation methods. We have extended this concept in our research to introduce a compu- tational primer validation method based on Fuzzy-Adaptive Particle Swarm Optimization (FAPSO) for the class of "species-speci c" primer design. Considering the requirements of unique primer design as well as experimental needs of PCR, we propose an overall design of a possible tool that can be implemented to gen- erate unique primers. Moreover, to validate our design, we have built a prototype by manipulating existing tools and algorithms according to our overall design. Considering the requirement of using resultant primers of a possible tool directly to a "wet-lab" ex- periment, we introduce a thorough evaluation scheme to evaluate a primer design tool of this kind. Based on the evaluation results of the primers generated by our prototype for serotypes of Dengue, we have identi ed that primers generated by our method can be directly used for a wet-lab experiment as the quality is higher than the primers generated by pre- existing methods. 2015-06-12T00:00:00Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/3153 Title: Automated Drinking Action Recognition in videos Authors: Kariyapperuma, K.M.M. Abstract: Today, human action recognition has been becoming one of the famous research areas in Computer Vision. This interest is motivated by a wide set of applications including security applications, video surveillance, video indexing, robot learning, etc. In terms of action recognition, recognizing actions that involve objects is a di cult task in current action recognition approaches. This research is to recognize such actions in another way to improve the accuracy by combining object detection and human action de- tection. In order to recognize the drinking action, a robust descriptor has been created to represent the body action by combining motion descriptor which has been calculated from optical ow and appearance descriptor which has been calculated from still video frames. This descriptor has been used in the action classi cation process. After detecting the action, object detection is conducted to verify the drinking action. Video clips containing drinking actions which have been extracted from movies, internet and some data sets were used in the training, testing and evaluation phases. In the evaluation, datasets were tested separately with action detection and object detection to identify the e ect of the veri cation process. 2015-06-08T00:00:00Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2485 Title: Secure Routing for Mobile Ad Hoc Networks Authors: Chathuranga, T.N. Abstract: Mobile Ad hoc Network (MANET) is a kind of wireless network which is formed by mobile nodes communicating with each other through wireless link without having xed infrastructure. Each device in MANET is free to move independently in any direction and change its links to other devices frequency. Securing wireless ad hoc networks is a highly challenging issue. Absence of any central co-ordination mechanism and shared wireless medium makes MANET more vulnerable to digitalcyber-attacks than wired network there are a number of attacks that a ect MANET. Proposed security mechanism more suitable for a less resource MANET such as Wireless Sensor Networks (WSN).It prevents routing attacks by au- thenticating and validating integrity of routing messages such as route re- quest and route reply with having an asymmetric security mechanism. Shar- ing a secret master key among all node in the network, it is impossible to achieve an asymmetric security. Therefore hash based message authenti- cation code (HMAC) M with a random number n which used to generate HMAC, are used with each routing messages by attaching them with the message. HAMC generation function works on cipher feedback mode and it will loop the process up to n times for a given routing message and given random number n, and return M. This mechanism is used in both route request and reply message generation and veri cation of those routing mes- sages. Proposed security mechanism courses to a route discovery latency but saves the average power consumption during the route discovery. It is important to select a better range for random number since it courses to the cost of computation. 2014-05-20T00:00:00Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2481 Title: Obstacle Classification through Acoustic Echolocation: A wearable obstacle detection and classification system for visually impaired persons Authors: Sampath, D.Y.K. Abstract: A visually impaired parson comes across objects with different attributes in his navigational path, and all of those objects can t be considered as obstacles. One person s obstacle could become a landmark for another. Therefore, getting some insights about the features of the obstacles that come across will provide a significant impact on improving the navigational process of the visually impaired individuals. In this study, an innovative wearable obstacle classification system which extends the currently available obstacle detection approaches using sonar echolocation to extract useful information about the obstacles while doing a basic obstacle classification, has been developed. In order to increase the detection range and to improve user-friendliness, an innovative approach based on integration of electronics onto textiles and wearable spectacles has been studied. Optimum architecture to embed electronic equipment s and sensors to the textile has been studied. Finally, a smart clothing and spectacles prototypes including ultrasonic sensors, vibration motors, power supplies and a micro controller has been developed. The working principle of the system is based on three main functions: sensing the surrounding environment via sonar sensors as well as classifying the detected obstacles into some meaningful set of classes and guiding the user to navigate through the environment by avoiding obstacles while giving the feedback of the characteristics of the obstacles comes across. The accuracy of the above functionality of the system was evaluated in unfamiliar indoor environments by using visually impaired individuals where extensive evaluations were carried out to check the classification accuracy of the system in three main aspects: classification of standard indoor target types, ground level obstacle types and complex target types as well. The evaluation results together with the feedback of the visually impaired individuals has proven that the proposed wearable smart prototype systems could become united part of visually impaired people s lifestyle, and it could help them overcome navigation concerns seamlessly, if the few drawbacks of the system like power consumption and cumbersomeness could be addressed further. 2014-05-20T00:00:00Z