In digital image mosaicking

search of seam points is the key procedure. It is usually to be done based on certain criteria to select the seam points row by row. In one-dimensional seam point search algorithm

the horizontal artificial edges are more significant than the vertical. In this article

a two-dimensional seam point searching algorithm is suggested where the grey level differences in the vertical direction are concerned as well as in the horizontal direction. Algorithm for mosaics of multispectral images is also introduced. Experiments show that with the method proposed

the artificial edges in the mosaics can be smoothed effectively.