Erhan Okuyan

Developed a video visualization application for ANKA UAV GCS with extended capabilities, such as:
• Allowing multiple video streams from the UAV to be displayed in real-time.
• Supporting H264 encoded videos with/without MpegTS and Stanag-7023 encapsulations.
• Facilitating customizable video data and video output buffering for smooth and low latency video display.
• Overlaying telemetry data of the UAV onto the video with custom symbology (dials, indicators, etc.).
• Sustaining telemetry from multiple sources.
• Facilitating the replay of previously recorded video and telemetry data.
• Enabling video exports.
• Fostering image processing algorithms, such as tracking and motion detection.
• Promoting the inference of external deep learning models, such as naval vehicle classification or forest fire detection.
The application was integrated into payload control, system management, and stand-alone software and actively used as the flight and mission monitoring tool for the ANKA UAV family. I was the sole developer on this project.

Built a many-to-many network stream routing application for ANKA UAV GCSs with extended capabilities, including:
• Allowing network data distribution to receive data from many network sources via UDP or TCP, then route the data to numerous destinations via UDP, TCP or even export the contents into a file.
• Replacing a single worker process created for each data source by implementing a multi-process architecture and improving the performance and reliability.
• Facilitating unicast and multicast UDP connections and client and server TCP options.
• Supporting multiple Ethernet interfaces.
• Allowing for optionally stripping Stanag7023 headers before routing the data.
• Permitting numerous running methods, whether via a windows service, a background application, or a Windows application with a GUI.
Ground control stations and the networks within these stations used the tool. I developed the entire application.

Operated as the sole developer to develop the graphical flight simulation application. EOFSim was designed to replace Flight Gear to visualize flight simulations for the ANKA UAV Family. It is an original implementation that does not use any high-level external libraries. EOFSim employed two graphics engines, one based on OpenGL and another on Direct3D. EOFSim has an object-oriented design. It used Digital Terrain Elevation Data (DTED) for terrain generation and uses AC models for airplane geometry. It utilized a dynamic view-dependent level-of-detail assignment mechanism and was multi-core optimized. EOFSim was made highly efficient, reaching 50+ fps on low-end computers using technologies such as C++, OpenGL 3.3, GLSL 3.3, Direct3D 9, and HLSL.

Created a visualization tool for unstructured volume data. The tool was developed as a part of my PhD studies. We proposed a new data representation for volume data, that allowed selective volume refinement. Afterward, we proposed a view-dependent refinement algorithm that utilized the volume data representation. Depending on the dataset, we were able to reduce the number of tetrahedra to render down to one-tenth of the original number with very little to no quality loss using this approach, which significantly reduced the rendering times.

Made a visualization tool for rendering unstructured volume data with GPU and multi-core CPUs. I developed this tool as a part of my PhD studies. We proposed parallelizing methods, a CPU-based Cell-Projection volume rendering algorithm for multi-core CPUs and the GPU. For multi-core parallelization, we managed to obtain almost linear speed-ups. For GPU parallelization, we managed to get practically interactable rendering times even with a low-end GPU achieving 20-30 times speed-ups.

Built a material visualization tool for solid-state physicists. The tool was created as part of my PhD studies. We collaborated with solid-state physics professors and developed this tool according to their needs. Using volume and surface rendering techniques, we visualized the structures and the defects of the solid materials.

Established a tool for pattern information extraction and defect quantification from crystal structures to assist crystallographers. This tool was created as a part of my PhD studies. We collaborated with solid-state physics professors and developed this tool according to their needs. We analyzed the crystal structures for local and global symmetries in the presence of defects to extract crystal parameters. We also updated the tool and published two follow-up papers in the same journal.