Region Extraction in Large-Scale Urban LIDAR Data

17
Region Extraction in Large-Scale Urban LIDAR Data Topics in Machine Vision Spring 2011 Alexandru Tandrau alexandru@tandrau .com

description

Seminar Presentation @Jacobs University, 02 March 2011. Based on A. Zavodny, P. Flynn, X. Chen. Region Extraction in Large-Scale Urban LIDAR Data. In 2009 IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops, pages 1801-1808.

Transcript of Region Extraction in Large-Scale Urban LIDAR Data

Page 1: Region Extraction in Large-Scale Urban LIDAR Data

Region Extraction in Large-Scale Urban LIDAR Data

Topics in Machine VisionSpring 2011

Alexandru [email protected]

Page 2: Region Extraction in Large-Scale Urban LIDAR Data

Input: Large 3D point cloud of urban environment, acquired by moving vehicles equipped with SICK LIDAR scanners, GPS-registered.

Output: Point cloud labeling by identifying planar regions

Page 3: Region Extraction in Large-Scale Urban LIDAR Data

NAVTEQ

Page 4: Region Extraction in Large-Scale Urban LIDAR Data

Beauvais Cathedral

Page 5: Region Extraction in Large-Scale Urban LIDAR Data

Algorithm Overview

1. Compute planarity measures2. Grow regions3. Combine regions4. Prune regions

Page 6: Region Extraction in Large-Scale Urban LIDAR Data

Best Fitting Plane

• Covariance Matrix gist

• Compute Eigenvalues• Compute Eigenvector v for smallest eigenvalue• Equation of a plane• Identify arguments of plane equation with

eigenvector gist

Tip The GSL (Gnu Scientific Library) contains, among others, eigenvalue/eigenvector libraries for C++.

Page 7: Region Extraction in Large-Scale Urban LIDAR Data

Compute Planarity Measures

Grow Regions

Combine Regions

Prune Regions

PM(Ps) =distToPlane(Pi,F)i

P

∑P

Approach 1

gist

gist

Page 8: Region Extraction in Large-Scale Urban LIDAR Data

Compute Planarity Measures

Grow Regions

Combine Regions

Prune Regions

Approach 2

Navg =N ii

N

∑N

PM(Ps) =angleBetween(Navg,N i)

2

i

N

∑N

Page 9: Region Extraction in Large-Scale Urban LIDAR Data

• ε-neighborhood with maximum planarity

– Find points to add to the region

– Update region’s planar approximation

• repeat for other neighborhoods

Compute Planarity Measures

Grow Regions

Combine Regions

Prune Regions

Page 10: Region Extraction in Large-Scale Urban LIDAR Data

• Regions Ri and Rj describe same surface if:– Normals of regions point in similar directions (Tnra)– The regions are close together

• Identify connected components on the graph formed by valid region pairs

Compute Planarity Measures

Grow Regions

Combine Regions

Prune Regions

d(Ri,R j ) = min∀p i ∈R i ,p j ∈R j

(dist(pi, p j ))

Page 11: Region Extraction in Large-Scale Urban LIDAR Data

Compute Planarity Measures

Grow Regions

Combine Regions

Prune Regions

Prune regions whose average planarity measure is above a threshold Tapm.

Page 12: Region Extraction in Large-Scale Urban LIDAR Data

Thresholds

• Neighborhood definitions– Epsilon– K

• Growing Regions– Trd

– Tpd

• Combining Regions– Tnra

– Tnrd

• Prunning Regions– Tapm

Page 13: Region Extraction in Large-Scale Urban LIDAR Data

Data Acquisition

point = (latitude, longitude, elevation)

Page 14: Region Extraction in Large-Scale Urban LIDAR Data

Experimental Results

Campus dataset, 95 mil. range-scanned points

Split data into subsets (based on acquire timestamp), process subsets independently

Execution time for the entire dataset21.4 minutes, ± 80% accuracy.

Page 15: Region Extraction in Large-Scale Urban LIDAR Data

A similar algorithm

[5]

Page 16: Region Extraction in Large-Scale Urban LIDAR Data

?

Presentation and collection of links also available at http://www.tandrau.com/mv_seminar

Page 17: Region Extraction in Large-Scale Urban LIDAR Data

References1. A. Zavodny, P. Flynn, X. Chen. Region Extraction in Large-Scale Urban LIDAR Data. In 2009

IEEE 12th International Conference on Computer Vision Workshops, ICCV Workshops, pages 1801-1808.

2. P. Flynn, A. Jain. Surface classification: hypothesis testing and parameter estimation. In Proceedings CVPR ‘88, pages 261-267, Jun 1988.

3. C. C. Chen and I. Stamos. Range image segmentation for modeling and object detection in urban scenes. In Proc. 3DIM ’07, pages 185-192, 2007.

4. P. K. Allen, A. Troccoli, B. Smith, I. Stamos, S. Murray. The Beauvais cathedral project. In Proc. Computer Vision and Pattern Recognition Workshop CVPRW ’03, volume 1, pages 10-10, June 2003.

5. J. Poppinga, N. Vaskevicius, A. Birk, K. Pathak. Fast Plane Detection and Polygonalization in noisy 3D Range Images. In 2008 IROS International Conference on Intelligent Robots and Systems.