Thesis supervisor: Levente Hajder
Location of studies (in Hungarian): ELTE Faculty of Informatics Abbreviation of location of studies: ELTE
Description of the research topic:
Over the last decade, 3D Computer Vision has become one of the most monumental fields of research. This large growth is mainly due to the development of autonomous technologies that is in the center of computer industry nowadays.
Applicants are tasked with the research and development of real/time algorithms using RGB cameras mounted on vehicles (in literature, this is known as visual odometry).
The focus of the research is the application of Affine Correspondences (Acs), they are the extension of Point Correspondences (PCs) and contain significantly more information with respect to a single PC. Currently, the exploits of PCs dominate computer vision literature. It is also a goal of this project to change that.
The project mainly consists of processing the data from vehicle mounted cameras, therefore the nature of such vehicles can be taken into consideration, e.g. the permanent connection of the ground and vehicle. Odometry for special vehicles, e.g. motorcycles, can also be researched.
Required language skills: English Further requirements: Strong mathematical background (linear algebra, estimation theory)