HEp-2 cell classification in indirect immunofluorescence image
ICICS'09 Proceedings of the 7th international conference on Information, communications and signal processing
| Hi-index | 0.00 |
The watershed transform is a popular segmentation method coming from the field of mathematical morphology. In order to solve the over-segmentation of watershed algorithm's applied to medical images, we first transform the original image into a morphological gradient image by using morphology methods and treat it with an opening-closing filter bank to get a reference image with less noise interference; then we segment the reference image by using the watershed algorithm; finally, according to the test criterion with respect to the gray mean of the region boundary and the region area, we merge the segment field. Experiments applying to segmentation of medical images by using watershed algorithm show that the image processing before and after segmentation prevents the over-segmentation effectively.