An Image Mosaic Algorithm Taking into Account Speed and Robustness

Abstract:

In this paper, a new image mosaic algorithm is proposed by studying the advantages and disadvantages of existing image mosaic algorithm. Firstly, finding a certain column at last frame in the two images which to be mosaic, making the gray difference between this column and other ones most obvious, that is to find the column whose feature is the most obvious as the matching template. Secondly, the correlated calculation is conducted between this column and each one at the next frame in turn,then the column that is the most relevant to the image at the next column is found and it is the best match one of the two images, thus the image mosaic is completed. Experiments show that the image mosaic can be successfully carried out by means of the algorithm.

• Periodical: Advanced Materials Research (Volume 267)
• Pages: 873-878
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.267.873
• Online since: June 2011
• Authors: Li Juan Zhang, Jin Hua Yang, Xiao Kun Wang
• Keywords: Image Fusion, Image Registration, Pearson Correlation Coefficient, Robustness

References

[1] David Peter Capel, in: Image Mosaicing and Super – resolution(University of Oxford Doctor Paper, Enland 2001).

[2] Zhong Li, Hu Xiaofeng, in: Overlapping image mosaic algorithm. edited by Journal of Image and Graphics, China(1998), pp.367-370.

[3] Sun Huayan, Zhou Daobing, Li Shengliang, in: A mosaic image algorithm for a sequence images, edited by Optics and Precision Engineering, China (2000), pp.35-37.

[4] G Stein, in: Accurate internal camera calibration using rota2tion with analysis of sources of error, edited by Proceedings of the 5th International Conference on Computer Vision, Cambridge MA(1995), pp.230-236.

DOI: 10.1109/iccv.1995.466781

[5] Richard Szeliski, in: Video mosaics for virtual environments, edited by IEEE Computer Graphics and Applications (1996), pp.222-230.