Research projects

Multimodal Transport Data Analytics Platform

Currently working as a researcher to develop a scalable, multimodal transport data aggregation and analytics platform which is capable of identifying emerging behaviors of passenger interests and demands, scheduling practices, driving behavior and habits, and monitoring and tracking fleets.

Integrated Flood Control and Water Management System

Currently working as a member in the research team to develop a center for flood controlling and water management.

Synthetic Aperture Radar (SAR) related Research work

• SAR reconstruction algorithm Optimizations on Knights Landing (C++).
• Energy aware SAR reconstruction on Hybrid multi-core systems (CUDA, C)
• Parallelization of Backprojection algorithm (CUDA, C)
• Change detection in SAR images using heterogeneous systems (CUDA, C)
• 2D and 3D SAR Simulator (C, Matlab)

GPU implementation of CMT-Nek

Was a research student at the Center for Compressible Multiphase Turbulence (CCMT) of University of Florida. Worked on the GPU implementation of CMT-Nek application.

Video Summarization methods using human computer interaction

Worked as a member in Human centered computing research group with Dr. Eakta Jain (UF). This research is about creating summaries for videos using data collected from human such as ECG and eye movements. We studied the behavior of heart rate while watching different kind of emotional scenes and came up with a relation between heart rate variation and the scene.

Other Projects

• RPAL parser and interpreter for Programming Language Principles course in UF ( C++)
• Ray Tracer for the project for Computer Graphics course in UF (C++)
• MIPS simulator for the final project for CDA5155 in UF (C++)
• Algorithm for Pattern Matching in Stars for Computer Geometry class in UF (C++)
• Web based data mining and analyzing system for IOM Sri Lanka (Java,GATE)
• Voice recognition system – Built a voice recognition system using Matlab software, which can train a model for a given voice and later can recognize the same voice. Machine learning methods such as KMeans algorithm are used to develop the software.
• Forensic Video and Image analyzing Tool Kit – For the final year research project developed a tool kit for analyzing forensic videos and images based on the research that was done. Methods were implemented to investigate different kind of forgeries in videos and images. Characteristics and properties of videos and video encoding mechanisms are used for the investigation. In addition to that methods were developed to extract maximum information from a given video by using fusion methods
• Camera driven robot with the ability of identifying road signals and travel – Built a robot which can identify road signals and response to them when travelling. ARM board is used to implement computer vision functionalities in LINUX operating system. 
• Intelligent Traffic light system – Developed methods to calculate the length of the queue in the road near to the traffic light system and depend on the length waiting times in the traffic is changed. Computer vision based techniques are used to calculate the amount of vehicles in the queue from the input video.
• OCR system for Sahana disaster management system – Designed mechanisms to correct the rotations in the data entry forms. Trained an OCR system to identify Sinhala language letters which is the national language in Sri Lanka.