Stitching Algorithm for Pathologic Slice Image Based on Wavelet Transform and Mathematical Morphology

摘要

In order to improve the stitching precision of pathologic slices under noise circumstances, a stitching algorithm based on wavelet transform and morphology is put forward in this paper. The maximal character object, which is closest to barycenter of overlapping region, is used to stitch images. Results show that the precision of registration is increased largely and the effect of mosaic is excellent with the method, and that it is a rapid and practical algorithm of image mosaic, especially applicable to image stitching under noise circumstances.