© The Institution of Engineering and Technology
Three main problems affect the alignment quality of existing studies on multi-resolution image stitching: (i) the initial motion obtained is sometimes incorrect; (ii) the local motion is hard to be estimated and (iii) the widely used global bundle adjustment is difficult to converge. The authors propose a new multi-resolution image mosaic method that combines three corresponding tactics to solve these problems. The first problem is solved by introducing an additional motion refinement strategy, which consists of the low-contrast filter and RANSAC. The former removes flatly textured surface pixels and thus eliminates the falsely matched features. The latter removes outliers and finds a robust initial motion for the next layer. The second problem is resolved by a new iteratively local registration method, which calibrates the current camera parameters based on those from previous image with robust non-linear optimisation methods. It improves the convergence efficiency and eliminates error minimisation. For the last problem, the authors introduce a five-parameter bundle adjustment method based on the axis-angle decomposition of the rotation matrix. Comparing with existing bundle adjustment methods, this method is more stable because of an accurate and simple rotation decomposition. The authors show the efficiency of the method with qualitative and quantitative experiments.
References
-
-
1)
-
D.G. Lowe
.
Distinctive image features from scale-invariant keypoints.
Int. J. Comput. Vis.
,
2 ,
91 -
110
-
2)
-
A. Noble
.
(1989)
Descriptions of image surfaces.
-
3)
-
R. Szeliski
.
Video mosaics for virtual environments.
IEEE Comput. Graph. Appl.
,
2 ,
22 -
30
-
4)
-
Harris, C., Stephens, M.: `A combined corner and edge detection', Proc. Fourth Alvey Vision Conf., 1988, p. 147–151.
-
5)
-
R. Szeliski
.
Image alignment and stitching: a tutorial.
Foundations Trends Comput. Graph. Comput. Vis.
,
1 ,
1 -
104
-
6)
-
M. Brown ,
D.G. Lowe
.
Automatic panoramic image stitching using invariant features.
Int. J. Comput. Vis.
,
1 ,
59 -
73
-
7)
-
Chen, T., Huang, T.S.: `Optimizing image registration by mutually exclusive scale components', IEEE 11th Int. Conf. on Computer Vision (ICCV 2007), 2007, p. 1–8.
-
8)
-
P.D. Kovesi
.
MATLAB and Octave functions for computer vision and image processing.
-
9)
-
X. Fang ,
M. Zhang ,
Z. Pan ,
P. Wang
.
A new method of manifold mosaic for large displacement images.
J. Comput. Sci. Technol.
,
2 ,
218 -
223
-
10)
-
T.F. Chan ,
S. Esedoglu
.
Aspects of total variation regularized ll function approximation.
SIAM J. Appl. Math.
,
5 ,
1817 -
1837
-
11)
-
J.J. Moré
.
(1978)
The Levenberg–Marquardt algorithm: implementation and theory’.
-
12)
-
X. Fang ,
Z. Pan ,
B. Luo ,
F. Wu ,
S. Guo
.
Robust image mosaic with RANSAC and bundle adjustment.
J. Comput. Inf. Syst.
,
4 ,
1613 -
1619
-
13)
-
Peleg, S., Herman, J.: `Panoramic mosaics by manifold projection', Proc. 1997 Conf. on Computer Vision and Pattern Recognition (CVPR 97), 1997, p. 338.
-
14)
-
K. Mikolajzyk ,
C. Schmid
.
A performance evaluation of local descriptors.
IEEE Trans. Pattern Anal. Mach. Intell.
,
10 ,
1615 -
1630
-
15)
-
Szeliski, R., Shum, H.Y.: `Creating full view panoramic image mosaics and environment maps', Proc. 24th Annual Conf. on Computer Graphics and Interactive Techniques (SIGGRAPH 97), 1997, p. 251–258.
-
16)
-
P.F. McLauchlan ,
A. Jaenicke
.
Image mosaicing using sequential bundle adjustment.
Image Vis. Comput.
,
751 -
759
-
17)
-
Cho, S.H., Chung, Y.K., Lee, J.Y.: `Automatic image mosaic system using image feature detection and taylor series', Proc. Seventh Int. Conf. on Digital Image Computing, Techniques and Applications (DICTA 2003), 2003, p. 549–560.
-
18)
-
Zhou, L.: `Matching using resolution pyramids with geometric constraints', Patent number US 6785427, 2004.
-
19)
-
R. Hartley
.
(2003)
Multiple view geometry in computer vision.
-
20)
-
Brown, M., Lowe, D.G.: `Recognising panoramas', ICCV '03: Proc. Ninth IEEE Int. Conf. on Computer Vision, 2003, p. 1218–1225.
-
21)
-
Bergen, J.R., Anandan, P., Hanna, K.J., Hingorani, R.: `Hierarchical model-based motion estimation', Proc. Second European Conf. on Computer Vision (ECCV'92), 1992, p. 237–252.
-
22)
-
T.F. Coleman ,
Y. Li
.
An interior trust region approach for nonlinear minimization subject to bounds.
SIAM J. Optim.
,
418 -
445
-
23)
-
Wang, C., Mahadevan, S.: `Manifold alignment using Procrustes analysis', Proc. 25th Int. Conf. on Machine Learning (ICM08), 2008, p. 1120–1127.
-
24)
-
H.S. Sawhney ,
R. Kumar
.
True multi-image alignment and its application to mosaicing and lens distortion correction.
IEEE Trans. Pattern Anal. Mach. Intell.
,
3 ,
235 -
243
-
25)
-
Y.-W. Huang ,
C.-Y. Chen ,
C.-H. Tsai ,
C.-F. Shen ,
L.-G. Chen
.
Survey on block matching motion estimation algorithms and architectures with new results.
J. VLSI Signal Process.
,
3 ,
297 -
320
-
26)
-
Bartoli, A., Zisserman, A.: `Direct estimation of non-rigid registration', Proc. British Machine Vision Conf. 2004 (BMVC 2004), 2004, p. 899–908.
-
27)
-
Shum, H.Y., Szeliski, R.: `Construction and refinement of panoramic mosaics with global and local alignment', Proc. Sixth Int. Conf. on Computer Vision (ICCV 98), 1998, p. 953.
-
28)
-
M.A. Fischler ,
R.C. Bolles
.
Random sample consensus: a paradigm formodel fitting with applications to image analysis and automated cartography.
Commun. ACM
,
6 ,
381 -
395
-
29)
-
B. Zitova ,
J. Flusser
.
Image registration methods: a survey.
Image Vis. Comput.
,
11 ,
977 -
1000
-
30)
-
A.R. Klumpp
.
Singularity-free extraction of a quaternion from a direction-cosine matrix.
J. Spacecraft Rockets
,
754 -
755
-
31)
-
C.L. Jeffrey ,
A.R. James ,
H.W. Margaret ,
E.W. Paul ,
E.W. Paul
.
Convergence properties of the Nelder–Mead simplex method in low dimensions.
SIAM J. Optim.
,
1 ,
112 -
147
http://iet.metastore.ingenta.com/content/journals/10.1049/iet-cvi.2009.0025
Related content
content/journals/10.1049/iet-cvi.2009.0025
pub_keyword,iet_inspecKeyword,pub_concept
6
6