UCSC Digital Library Collection:
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4139
2023-09-26T18:29:28ZMedical Prescription Identification Solution
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4211
Title: Medical Prescription Identification Solution
Authors: Wijewardena, W.R.A.D.
Abstract: A medical prescription is a very familier document to any person and is a one that
is usually believed impossible to read. There are several reasons for that well known
conclustion. First, the sloppy handwriting of the doctors and second, the lack of
the domain knowledge. Because of these difficulties in reading a prescription have
a high probability of ending up in misreading the content. These misreading often
lead to many health issues with regard to the patient and even a threat to their
lives. But unfortunatly both the mentioned reasons for such situations cannot be
changed.
In the present, one of the leading research area is Optical Character Recognition. Among that, handwritten character recognition takes a significant interest in
researchers. Taking these advantaged into account and with the help of the domain knowledge, this research is to find a way to accuratly read the content of a
medical prescription. This research uses a neural network approach for the charater
recognition process and a knowledge base matching to accuralty output the result.
The outcome of this research has been a successfull enhancement in the prescription
identification domain and has established for further improvements.2021-07-26T00:00:00ZEfficiently transform contracts written in Peyton Jones contract descriptive language to Solidity
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4210
Title: Efficiently transform contracts written in Peyton Jones contract descriptive language to Solidity
Authors: Wickramarachchi, V. U.
Abstract: Financial contracts play a major role in the modern economy. Due to a large variation of
contracts being traded in financial markets, a standard representation for contracts was
required in order to eliminate the ambiguity imposed by natural languages. Peyton Jones et
al. catered this need by proposing a combinator library embedded in Haskell which enabled
contract representation and valuation. However, every financial contract represented as such
suffered from the same fundamental problem. If a contract is executable, the parties
involved in the contract needed to trust a central counterparty to give them the correct results
after execution. The interference of this middleman introduced certain risks as well as a
significant amount of overhead.
In this dissertation, a novel approach to facilitate autonomous contract execution was
proposed, exploiting the features and use cases of the Ethereum blockchain and its scripting
language, Solidity. The approach involves transforming contracts written in the Peyton
Jones’ Contract Descriptive language to Solidity using a special purpose compiler. The
result of this transformation is a smart contract equivalent to the traditional financial
contract. The transformed smart contract is deployed and executed on the Ethereum
blockchain using an Ethereum client.
The proposed solution was evaluated against existing attempts to design autonomous
financial contracts. The research shows that a contract written in the Peyton Jones’ Contract
Descriptive language can be transformed to a smart contract which executes autonomously
in a trustless environment. As a result, it was concluded that financial contracts could
operate without a central counterparty with increased performance and reduced overheads.2021-07-26T00:00:00ZCombining Image Processing Techniques and Mobile Sensor Information for Marker-less Augmented Reality Based Reconstruction
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4209
Title: Combining Image Processing Techniques and Mobile Sensor Information for Marker-less Augmented Reality Based Reconstruction
Authors: Weerakkody, I. S.
Abstract: Marker-less augmented reality based reconstruction using mobile devices, is a near
impossible task. This largely due to the lack of processing power in mobile devices
when considering vision based tracking for localization and to the lack of accuracy in
mobile GPS when considering mobile sensor based approach for localization.
In order to address this problem this research presents a novel approach which
combines image processing techniques and mobile sensor information which can be
used to perform precise position localization in order to perform augmented reality
based reconstruction using mobile devices. The core of this proposed methodology is
tightly bound with the image processing technique which is used to identify the object
scale in a given user image. Use of mobile sensor information was to classify the most
optimal locations for a given particular user location.
This proposed methodology has been evaluated against the results obtained using
10cm accurate RTK device and against the results obtained using only the A-GPS
chips in mobile device. Though this proposed methodology require more processing
time than A-GPS chips, the accuracy level of this proposed methodology is
outperforms that of A-GPS chips. And the results of the experiments carried out
further convince that this proposed methodology facilitates improving the accuracy of
position localization for augmented reality based reconstruction using mobile devices.2021-07-26T00:00:00ZA Blockchain Based Approach for Secure E-Voting System
https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4208
Title: A Blockchain Based Approach for Secure E-Voting System
Authors: Sanjaya, M.D.
Abstract: Voting is the process of representation of democracy in a country, to select a person
for parliament, approve a bill in a parliament and decision making in board meetings
etc. Voting systems have been around of hundreds of years and but they were evolving
very slowly. Many solutions were proposed in the history, but most of them were
rejected because of some security issues and limitations. Finally at the 21th century, evoting systems started to bloom with the development of the web technologies
With the development of the blockchain 2.0, the researchers started to go towards a
new destination by applying blockchain to software engineering applications. E-voting
systems were developed based on Ethereum as well as Zcash and bitcoin. But they
were not full e-voting frameworks. Due to the limitations of proposed solution, those
were unable to exist with the modern world.
In order to prove feasibility of developing a secure e-voting system by combining
some concepts of Zcash with the ethereum platform, this research is a proof of concept
to implement a full e-voting framework with voter registration, voter verification,
voting, tallying and end to end verification. To maintain user privacy, zk-SNARK
which is a concept used in Zcash for maintaining private transactions, was used. To
write Immutable codes, a concept called smart contract which is used in ethereum, was
used. Due to the experiment and testing done, it is very clear that this system is a
practical solution and works well by protecting voter privacy.2021-07-26T00:00:00Z