In complex terrains or non-airport environments, the takeoff and landing of a helicopter is often challenged by factors such as space limitation, terrain complexity, and surrounding obstacles, which demands a great deal on pilots' perception and decision-making. In order to face these challenges, this thesis created a helicopter landing assistance system based on multi-source geographical data. This system integrates various geodata, including the Digital Surface Model, the orthophoto, LiDAR point clouds, and so on, to construct a 3D scenario to simulate the flight process and provide extra information for pilots. Through plane fitting and clustering, we can analyze the obstacles in the potential landing site, and then classify them using a random forest classifier. Finally, these data is combined with a terrain model in Unity to visualize an interactive environment, which is later synchronized with an actual helicopter cockpit simulator. The entire system was eventually evaluated through demonstration videos and a survey questionnaire, which showed positive effects in data integration, obstacle detection, and decision-making support. Overall, this thesis proves the potential of supporting helicopter flight and landing through the integration and analysis of multi-source geographical data.

Procedural Content Generation (PCG) reduces the manual effort required to create
content and increases variety, yet creating endless urban worlds that remain legible and
engaging is challenging. This thesis investigates the generation of infinite, traversable
city spaces that support navigation and meaningful player experience. A city generator
was developed using the Wave Function Collapse (WFC) algorithm, extended with
three mechanisms: weighted tile selection, vicinity-based bans to reduce repetition,
and a point of interest (POI) extension that integrates landmarks as functional anchors.
The system was evaluated for structural quality, variety, and runtime performance.
Weighted and vicinity-ban WFC reduced repetitive road patterns, while POIs offer
orientation and possible encouragement for exploration. Performance analysis showed
that propagation dominated runtime costs but remained within practical limits for
real-time use. Limitations were identified in road topology, biome integration, and
long-term variety. Overall, the results demonstrate how practical adjustments to WFC
and a POI-based framing can balance algorithmic control with navigability and player
experience in procedural city generation.

Mixed Reality (MR) applications depend on real-time data exchange, yet current
publish–subscribe frameworks like Ubi-Interact (UBII) rely solely on topic-based routing.
This context-agnostic approach delivers all topic data to all subscribers, leading to
wasted bandwidth, unnecessary device processing, and potential privacy risks.
This thesis proposes a context-enabled publish–subscribe model for MR environ-
ments, extending UBII with location-aware mechanisms. Protocol Buffer extensions
support static and dynamic location references, while backend services implement a
LocationManager to process Global Positioning System (GPS) updates, compute Haver-
sine distances, and apply spatial filtering. A web-based frontend integrates browser
geolocation for testing and development.
Evaluation shows that location-aware routing reduces irrelevant message delivery
and improves efficiency without compromising real-time responsiveness. Privacy
concerns are addressed through configurable precision and obfuscation strategies. The
work concludes by outlining future directions for hybrid context models, adaptive
filtering, and privacy-preserving policies, paving the way for efficient and secure MR
middleware.

Augmented Reality (AR) is emerging as one of the most transformative technological trends since the introduction of the smartphone. It has the potential to fundamentally change the way we access and experience digital content by seamlessly integrating virtual elements into our physical environment. In recent years, significant advancements in display technology, spatial computing, sensor integration, and wireless connectivity have accelerated the development of AR glasses, bringing the vision of immersive, always-available digital interfaces closer to reality. Despite growing success in enterprise environments, AR has yet to gain meaningful traction among everyday consumers. This thesis provides a comprehensive overview of the current state of AR technology, critically examines its potential as a next-generation computing platform, and explores the technological and socio-cultural challenges that must be overcome for widespread consumer adoption to occur. Special attention is given to the differences between industrial and consumer requirements, the limitations of current devices, and the essential features future AR systems must offer in order to become a ubiquitous part of daily life.