Using 3D GNG-based reconstruction for 6DoF egomotion

Viejo, Diego; García, José Luis Simón; Cazorla, Miguel; Gil, David; Johnsson, Magnus

doi:10.1109/ijcnn.2011.6033337

Cited by 2 publications

(6 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This registration can provide a good starting point for Simultaneous Location and Mapping (SLAM). We use the method proposed in [31]. This method is developed for managing 3D point sets collected by any kind of sensor.…”

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

“…This method is developed for managing 3D point sets collected by any kind of sensor. For our experiments, we have used data from an infrared time-of-flight camera SR4000, but in [31] there are examples of this method applied to other 3D devices, like a sweeping unit with a 2D laser Sick and a Digiclops stereo camera, mounted on a mobile robot. We are also interested in dealing with outliers, i.e., environments with people or non-modeled objects.…”

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

“…We briefly describe the method proposed in [31] to manage 3D data and to use it for 6DoF egomotion calculation. GNG produces a Delaunay Triangulation which can be used as a representation of the points neighbourhood.…”

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

“…To validate our work we have applied our GNG accelerated implementation to the extraction and model of features from 3D raw data [31,32,33,34]. Moreover, using this method, apart from accelerating the routine, we achieve two other advantages: a complexity reduction (when comparing with raw data) and an improvement of speed-up without decreasing the quality of the representation obtained.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

GPGPU implementation of growing neural gas: Application to 3D scene reconstruction

Orts

García-Rodríguez

Viejo

et al. 2012

Journal of Parallel and Distributed Computing

Self Cite

View full text Add to dashboard Cite

Self-organising neural models have the ability to provide a good representation of the input space. In particular the Growing Neural Gas (GNG) is a suitable model because of its flexibility, rapid adaptation and excellent quality of representation.However, this type of learning is time consuming, specially for high-dimensional input data. Since real applications often work under time constraints, it is necessary to adapt the learning process in order to complete it in a predefined time.This paper proposes a Graphics Processing Unit (GPU) parallel implementation of the GNG with Compute Unified Device Architecture (CUDA). In contrast to existing algorithms, the proposed GPU implementation allows the acceleration of the learning process keeping a good quality of representation. Comparative experiments using iterative, parallel and hybrid implementation are carried out to demonstrate the effectiveness of CUDA implementation. The results show that GNG learning with proposed implementation achieves a speed-up of 6x comEmail addresses: sorts@dtic.ua.es (Sergio Orts), jgarcia@dtic.ua.es (Jose Garcia-Rodriguez), dviejo@dccia.ua.es (Diego Viejo), miguel.cazorla@ua.es (Miguel Cazorla), vmorell@dccia.ua.es (Vicente Morell) Preprint submitted to Parallel and Distributed Computing May 9, 2012 pared with single-threaded CPU implementation. GPU implementation has been also applied to a real application with time constraints: acceleration of 3D scene reconstruction for egomotion, in order to validate the proposal.

show abstract

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

Section: Accelerating 6dof Egomotion Using Gngmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

GPGPU implementation of growing neural gas: Application to 3D scene reconstruction

Orts

García-Rodríguez

Viejo

et al. 2012

Journal of Parallel and Distributed Computing

Self Cite

View full text Add to dashboard Cite

show abstract

“…In our previous work Viejo et al (2011) we proposed a method for extracting and modeling planar patches from 3D raw data. Using this method we achieve two main advantages: first, a complexity reduction (when comparing with raw data) is done and time and memory consumptions are improved (we obtain over 500 features from 100000 3D points); second, outliers are better overcome using these features, as points not sup-ported by a planar patch are deleted.…”

Section: Introductionmentioning

confidence: 99%

Using GNG to improve 3D feature extraction—Application to 6DoF egomotion

et al. 2012

Self Cite

View full text Add to dashboard Cite

Several recent works deal with 3D data in mobile robotic problems, e.g. mapping or egomotion. Data comes from any kind of sensor such as stereo vision systems, time of flight cameras or 3D lasers, providing a huge amount of unorganized 3D data. In this paper, we describe an efficient method to build complete 3D models from a Growing Neural Gas (GNG). The GNG is applied to the 3D raw data and it reduces both the subjacent error and the number of points, keeping the topology of the 3D data. The GNG output is then used in a 3D features extraction method. We have performed a deep study in which we quantitatively show that the use of GNG improves the 3D feature extraction method. We also show that our method can be applied to any kind of 3D data. The 3D features obtained are used as input in an Iterative Closest Point (ICP)-like method to compute the 6DoF movement performed by a mobile robot. A comparison with standard ICP is performed, showing that the use of GNG improves the results. Final results of 3D mapping from the egomotion calculated are also shown.

show abstract

Using 3D GNG-based reconstruction for 6DoF egomotion

Cited by 2 publications

References 14 publications

GPGPU implementation of growing neural gas: Application to 3D scene reconstruction

GPGPU implementation of growing neural gas: Application to 3D scene reconstruction

Using GNG to improve 3D feature extraction—Application to 6DoF egomotion

Contact Info

Product

Resources

About