Public defence in Computer Science, M.Sc. (Tech) Otto Korkalo

Title of the thesis: Systems and Methods for Multiple-View and Depth-Based People Tracking and Human-Computer Interaction

Doctoral student: Otto Korkalo
Opponent: Professor Laura Ruotsalainen, University of Helsinki
Custos: Emeritus Professor Tapio Takala, Aalto University School of Science, Department of Computer Science

As information technology systems become more and more integrated into everyday environments, users interact directly with spaces and objects. Estimating and analyzing people flow, intelligent control of building ventilation and lighting, interactive games in public spaces, and mixed reality assisted industrial assembly are examples of such applications. It is crucial to determine where and how people move, what they do, how spaces can be made interactive, and how virtuality can be integrated into the physical, three-dimensional world. The dissertation delved into developing machine vision-based solutions and methods for intelligent environments, human-computer interaction, and mixed reality. Real-time location and orientation data were generated by combining information from multiple cameras and utilizing depth cameras, which provide measurements of the environment directly in three dimensions. 

As part of the dissertation, a method for the automatic calibration of a human tracking system based on multiple depth cameras was developed, allowing for easier and faster deployment of such systems in practical installations. To implement interactive multi-touch displays, a multi-camera-based touch detection and tracking method was created. For augmented reality applications, a method was developed to accurately determine the position and orientation of depth cameras relative to the operating environment. 

In contrast to previous research, the calibration method developed for human tracking allows for the determination of camera positions, orientations, and measurement errors, enabling optimal integration of results from various cameras. The solution developed for multi-touch displays enables any flat display to be transformed into a multi-touch screen, seamlessly scaling to different installations. The camera tracking method developed for mixed reality facilitates a simpler and more precise fitting of virtual models into the physical world. 

The solutions developed in this work have been applied in numerous practical prototypes and public installations. These have served to demonstrate both the solutions and methods developed in the dissertation, as well as the broader opportunities provided by new interaction technologies as the boundary between physical and digital environments becomes increasingly blurred.

Thesis available for public display 10 days prior to the defence at: https://aaltodoc.aalto.fi/doc_public/eonly/riiputus/ 

Contact information:

Doctoral theses of the School of Science: https://aaltodoc.aalto.fi/handle/123456789/52