Learning and Understanding in 3D Point Clouds Scientific Director: Tolga Birdal, Slobodan Ilic Contact Person(s): Tolga Birdal In industrial collaboration with: Siemens AG

Keywords: Registration / Visualization, Reconstruction, Computer Vision, Sensing

Abstract

Point offer are a unified representation for 3D data. Whether the sensor is a laser scanner, lidar or a depth camera, the acquired information can be stored in form an unstructured point cloud. This makes such representation ubiquitous. In this project, we aim to adapt state-of-the-art learning algorithms to efficiently handle point cloud data in order to find objects, detect structures, perform segmentation and 3d reconstruction.

Videos

We present a revised pipe-line of the existing 3D object detection and pose estimation framework based on point pair feature matching. In particular, we propose to couple the object detection with a coarse-to-fine segmentation, where each segment is subject to disjoint pose estimation. During matching, we apply a weighted Hough voting and an interpolated recovery of pose parameters. Finally, all the generated hypotheses are tested via an occlusion-aware ranking and sorted. We argue that such a combined pipeline simultaneously boosts the detection rate and reduces the complexity, while improving the accuracy of the resulting pose.

Publications

Sorry! No entries found for your search query.

Team

Contact Person(s)

M.Sc. Tolga Birdal

Working Group

Dr. Slobodan Ilic

M.Sc. Tolga Birdal

Location

Technische Universität München Institut für Informatik / I16 Boltzmannstr. 3 85748 Garching bei München Tel.: +49 89 289-17058 Fax: +49 89 289-17059

Visit our lab at Garching.



internal project page

Please contact Tolga Birdal for available student projects within this research project.