GPU-Accelerated Nearest Neighbor Search for 3D Registration

Authors:
Deyuan Qiu;Stefan May;Andreas Nüchter
Affiliations:
University of Applied Sciences Bonn-Rhein-Sieg, Sankt Augustin, Germany;INRIA, Sophia-Antipolis, France;Jacobs University Bremen, Germany
Venue:
ICVS '09 Proceedings of the 7th International Conference on Computer Vision Systems: Computer Vision Systems
Year:
2009

Citing 9
Cited 2

A Method for Registration of 3-D Shapes

IEEE Transactions on Pattern Analysis and Machine Intelligence - Special issue on interpretation of 3-D scenes—part II
Realistic image synthesis using photon mapping

Realistic image synthesis using photon mapping
Photon mapping on programmable graphics hardware

Proceedings of the ACM SIGGRAPH/EUROGRAPHICS conference on Graphics hardware
KD-tree acceleration structures for a GPU raytracer

Proceedings of the ACM SIGGRAPH/EUROGRAPHICS conference on Graphics hardware
Interactive k-d tree GPU raytracing

Proceedings of the 2007 symposium on Interactive 3D graphics and games
Real-time KD-tree construction on graphics hardware

ACM SIGGRAPH Asia 2008 papers
SIMD Packet Techniques for Photon Mapping

RT '07 Proceedings of the 2007 IEEE Symposium on Interactive Ray Tracing
A Fast Similarity Join Algorithm Using Graphics Processing Units

ICDE '08 Proceedings of the 2008 IEEE 24th International Conference on Data Engineering
A graphics hardware accelerated algorithm for nearest neighbor search

ICCS'06 Proceedings of the 6th international conference on Computational Science - Volume Part IV

Unstructured point cloud surface denoising and decimation using distance RBF K-nearest neighbor Kernel

PCM'10 Proceedings of the Advances in multimedia information processing, and 11th Pacific Rim conference on Multimedia: Part II
kANN on the GPU with shifted sorting

EGGH-HPG'12 Proceedings of the Fourth ACM SIGGRAPH / Eurographics conference on High-Performance Graphics

Quantified Score

Hi-index	0.00

Visualization

Abstract

Nearest Neighbor Search (NNS) is employed by many computer vision algorithms. The computational complexity is large and constitutes a challenge for real-time capability. The basic problem is in rapidly processing a huge amount of data, which is often addressed by means of highly sophisticated search methods and parallelism. We show that NNS based vision algorithms like the Iterative Closest Points algorithm (ICP) can achieve real-time capability while preserving compact size and moderate energy consumption as it is needed in robotics and many other domains. The approach exploits the concept of general purpose computation on graphics processing units (GPGPU) and is compared to parallel processing on CPU. We apply this approach to the 3D scan registration problem, for which a speed-up factor of 88 compared to a sequential CPU implementation is reported.