Institute of Computer Graphics and Vision, TU Graz
The Institute for Computer Graphics and Vision (ICG) at Graz University of Technology, is the only Austrian academic group with the charter to address both computer vision and computer graphics. The ICG is carefully nurturing a culture of digital visual information processing.
Research at the ICG with Prof. Horst Bischof, Prof. Vincent Lepetit, Prof. Thomas Pock and Prof. Dieter Schmalstieg is focused on machine vision, machine learning, medical image analysis, object reconstruction and object recognition, computer graphics and visualization.
The institute is home to 19 civil service positions and about another 70 soft money positions, making it one of the largest institutes at Graz University of Technology. During the most recent 5-year period, the institute was responsible for 80 diploma theses and the award of 40 doctorates.
Over the last years, our researchers authored on average about 90 publications a year in scientific journals, book chapters and international conferences. The ICG has a lot of experience in national (FWF and FFG) as well as EU-funded projects.
2D object detection
3D object detection
lane marking detection
The Institute of Computer Graphics and Vision develops the algorithms for interpreting the LiDAR point clouds and assessing the semantic meaning of the acquired data. Know-how on all aspects of computer vision and the processing of depth information can be provided to best possibly utilize the new LiDAR sensor. Conventional computer vision as well as variational methods and deep learning methods are deployed to process the laser range data.
The aim is to get a comprehensive understanding of the scene around the vehicle which is essential to autonomous systems to safely navigate and avoid dangers. This includes the detection of other road users as well as the extraction of static information like lane markings.
LiDAR Processing, Object Detection and Classification (WP5)
The goal of this workpackage is to develop algorithms to process the LiDAR point clouds, from the data preparation to the final extraction of semantic information. This includes methods for pre-processing such as the reduction of acquisition noise. Based on the resulting point clouds, objects of interest such as cars or pedestrians are detected in the scene and segmented from the background. Besides localizing and segmenting objects, they are classified into different classes. Once objects have been found and delineated, they need to be tracked over time to estimate the movement of the object. Aside from the localization, classification and tracking of objects, lane markings are detected and tracked over time to estimate the lane-path.