This article proposes a new characteristic-based image mosaic algorithm that used the granular computing theory. Firstly, establish a granular computing model of the image waiting to be spliced to obtain the image's edge segmentation; Secondly, extract the characteristic points in the edge graph, then carry on the related operation to these characteristic points to find matching characteristic points, and eliminate the mismatches to find the most superior matching characteristics; Thirdly, work out the spot transformation relations of two images according to the matching characteristic points, and mapped the two images on the same geometric plane surface by transformation matrix; Finally, smooth transition processing will be carried on the overlapped region to achieve the purpose of seamless image mosaic. The experimental result indicates that the algorithm in this article is a stable and steady method that superior than the traditional algorithm, and it gets well effectiveness.
[1]
Tsau Young Lin,et al.
Granular computing: examples, intuitions and modeling
,
2005,
2005 IEEE International Conference on Granular Computing.
[2]
M. Mizumoto.
Pictorial representations of fuzzy connectives, part I: cases of t-norms, t-conorms and averaging operators
,
1989
.
[3]
Witold Pedrycz,et al.
Fuzzy relational compression
,
1999,
IEEE Trans. Syst. Man Cybern. Part B.
[4]
Liu,et al.
Formalization for Granular Computing Based on Logical Formulas
,
2006
.
[5]
Edward H. Adelson,et al.
A multiresolution spline with application to image mosaics
,
1983,
TOGS.