Thesis supervisor: Zoltán Rózsa
Location of studies (in Hungarian): Department of Material Handling and Logistics Systems Abbreviation of location of studies: ALRT
Description of the research topic:
Aim of the research: Robot vehicles today are equipped with several environmental perception sensors (camera, LIDAR, radar, etc.) to serve pilot support systems, remote control and self-driving. However, we are far from the optimal usage of these equipment from the point of view of processing their data. By combining the benefits of deep-learning networks and traditional geometric based methods, a new level of processing can be achieved, where processing can be faster, more accurate, or even new types of data can be available.
The tasks to be performed include:
• production of measurement data, sensors for that like camera, LIDAR, etc. are available;
• processing of sensor data, extraction of characteristics;
• development of new traditional and deep-learning methods
• by fusing them, speed up and accuracy of detection is aimed to achieve.
As a first step, it is necessary to expand the research in several directions to explore the application possibilities of different intelligent sensors, 3D imaging and camera systems. Next, it is required to examine traditional and deep-learning methods for utilization in 3D reconstruction, by focusing on static objects first, then dynamic ones. Finally, by fusing these, new methods should be developed to optimize the different parameters.
Expected scientific results: Optimized data processing (shape recognition, localization, mapping, etc.) can be able to greatly advance practical applications.
Required language skills: English Further requirements: High level MATLAB and / or C / C ++ programming knowledge required. Computer vision knowledge is an advantage.
Number of students who can be accepted: 2
Deadline for application: 2023-09-20
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).
Login
