›› 2012, Vol. 27 ›› Issue (6): 1269-1280.doi: 10.1007/s11390-012-1303-3

• Graphics, Visualization, and Image Processing • Previous Articles     Next Articles

Hybrid Parallel Bundle Adjustment for 3D Scene Reconstruction with Massive Points

Xin Liu (刘鑫), Wei Gao (高伟), and Zhan-Yi Hu (胡占义)   

  1. National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Sciences, Beijing 100190, China
  • Received:2011-10-11 Revised:2012-07-07 Online:2012-11-05 Published:2012-11-05
  • Supported by:

    This work was supported by the National Natural Science Foundation of China under Grant No. 60835003 and the Strategic Priority Research Program of the Chinese Academy of Sciences under Grant No. XDA06030300.

Bundle adjustment (BA) is a crucial but time consuming step in 3D reconstruction. In this paper, we intend to tackle a special class of BA problems where the reconstructed 3D points are much more numerous than the camera parameters, called Massive-Points BA (MPBA) problems. This is often the case when high-resolution images are used. We present a design and implementation of a new bundle adjustment algorithm for efficiently solving the MPBA problems. The use of hardware parallelism, the multi-core CPUs as well as GPUs, is explored. By careful memory-usage design, the graphic-memory limitation is effectively alleviated. Several modern acceleration strategies for bundle adjustment, such as the mixed-precision arithmetics, the embedded point iteration, and the preconditioned conjugate gradients, are explored and compared. By using several high-resolution image datasets, we generate a variety of MPBA problems, with which the performance of five bundle adjustment algorithms are evaluated. The experimental results show that our algorithm is up to 40 times faster than classical Sparse Bundle Adjustment, while maintaining comparable precision.

[1] Agarwal S, Snavely N, Simon I, Seitz S M, Szeliski R. Build-ing Rome in a day. In Proc. the 12th ICCV, Sept. 2001,pp.72-79.

[2] Frahm J M, Georgel P, Gallup D et al. Building Rome on acloudless day. In Proc. the 11th ECCV, Sept. 2010, pp.368-381.

[3] Snavely N, Seitz S, Szeliski R. Photo tourism: Exploringphoto collections in 3D. In Proc. SIGGRAPH, Jul. 30-Aug. 3,2006, pp.835-846.

[4] Agarwal S, Snavely N, Seitz S et al. Bundle adjustment in thelarge. In Proc. the 11th ECCV, Part 2, Sept. 2010, pp.29-42.

[5] Wu C C, Agarwal S, Curless B, Seitz S. Multicore bundleadjustment. In Proc. CVPR, June 2011, pp.3057-3064.

[6] Lourakis M, Argyros A. SBA: A software package for genericsparse bundle adjustment. ACM Transactions on Mathemat-ical Software, 2009, 36(1), Article No. 2.

[7] Triggs B, Mclauchlan P, Hartley R et al. Bundle adjustment— A modern synthesis. In Proc. the Int. Workshop onVision Algorithms: Theory and Practice, Sept. 1999, pp.298-372.

[8] Byröd M, Aström K. Bundle adjustment using conjugate gra-dients with multiscale preconditioning. In Proc. BMVC,Sept. 2009.

[9] Byröd M, oAström K. Conjugate gradient bundle adjustment.In Proc. the 11th ECCV, Sept. 2010, pp.114-127.

[10] Jeong Y, Nister D, Steedly D et al. Pushing the envelopeof modern methods for bundle adjustment. In Proc. CVPR,June 2010, pp.1474-1481.

[11] Choudhary C, Gupta S, Narayanan P J. Practical time bun-dle adjustment for 3D reconstruction on the GPU. In ECCVWorkshop on Computer Vision on GPUs, Sept. 2010.

[12] Chen D H, Chen W G, Zheng W M. CUDA-Zero: A frame-work for porting shared memory GPU applications to multi-GPUs. Science China Information Sciences, 2012, 55(3):663-676.

[13] Yang X, Tang T, Wang G et al. MPtostream: An OpenMPcompiler for CPU-GPU heterogeneous parallel systems. Sci-ence China Information Sciences, 2012, 55(9): 1961-1971.

[14] Engels C, Stewenius S, Nister D. Bundle adjustment rules. InProc. PCV, Sept. 2006, pp.266-271.

[15] Shonkwiler R, Lefto L. An Introduction to Parallel and VectorScientific Computing. London: Cambridge University Press,2006.

[16] Park I Y, Singhal N, Lee M H et al. Design and performanceevaluation of image processing algorithms on GPUs. IEEETrans. Parallel and Distributed Systems, 2011, 22(1): 91-104.

[17] Castano-Díez D, Moser D, Schoenegger A et al. Performanceevaluation of image processing algorithms on the GPU. Jour-nal of Structural Biology, 2008, 164(1): 153-160.

[18] Bordawekar R, Bondhugula U, Rao R. Can CPUs matchGPUs on performance with productivity?: Experiences withoptimizing a FLOP-intensive application on CPUs and GPU.Technical Report, IBM T. J. Watson Research Center, 2010.

[19] Alcantara D, Sharf A, Abbasinejad F et al. Real-time parallelhashing on the GPU. ACM Transactions on Graphics, 2009,28(5), Article No. 154.
No related articles found!
Full text



[1] Liu Mingye; Hong Enyu;. Some Covering Problems and Their Solutions in Automatic Logic Synthesis Systems[J]. , 1986, 1(2): 83 -92 .
[2] Chen Shihua;. On the Structure of (Weak) Inverses of an (Weakly) Invertible Finite Automaton[J]. , 1986, 1(3): 92 -100 .
[3] Gao Qingshi; Zhang Xiang; Yang Shufan; Chen Shuqing;. Vector Computer 757[J]. , 1986, 1(3): 1 -14 .
[4] Chen Zhaoxiong; Gao Qingshi;. A Substitution Based Model for the Implementation of PROLOG——The Design and Implementation of LPROLOG[J]. , 1986, 1(4): 17 -26 .
[5] Huang Heyan;. A Parallel Implementation Model of HPARLOG[J]. , 1986, 1(4): 27 -38 .
[6] Shen Li; Stephen Y.H.Su;. Generalized Parallel Signature Analyzers with External Exclusive-OR Gates[J]. , 1986, 1(4): 49 -61 .
[7] Min Yinghua; Han Zhide;. A Built-in Test Pattern Generator[J]. , 1986, 1(4): 62 -74 .
[8] Tang Tonggao; Zhao Zhaokeng;. Stack Method in Program Semantics[J]. , 1987, 2(1): 51 -63 .
[9] Min Yinghua;. Easy Test Generation PLAs[J]. , 1987, 2(1): 72 -80 .
[10] Zhu Hong;. Some Mathematical Properties of the Functional Programming Language FP[J]. , 1987, 2(3): 202 -216 .

ISSN 1000-9000(Print)

CN 11-2296/TP

Editorial Board
Author Guidelines
Journal of Computer Science and Technology
Institute of Computing Technology, Chinese Academy of Sciences
P.O. Box 2704, Beijing 100190 P.R. China
E-mail: jcst@ict.ac.cn
  Copyright ©2015 JCST, All Rights Reserved