Modification on distance transform to avoid over-segmentation and under-segmentation

摘要

Many techniques have been proposed in literature, both conventional and morphological, for the segmentation of images. Image segmentation using morphological methods, particularly using watershed algorithm has found wide applications. The popular Beucher & Meyer's method based on marker constrained watershed algorithm for image segmentation, which uses only the binary image and its distance transform, can lead to oversegmentation or under-segmentation if the image objects are of irregular shape and the objects are overlapping or touching. Another problem with the Beucher & Meyer's method is that the watersheds constructed may not agree/coincide with the actual image object junctions as only the binary image is used and information contained in the gray scale image is not utilised. In this paper we are presenting an approach addressing both these problems, using morphological area-opening on the distance transform for choosing markers to avoid over-segmentation as well as under-segmentation and the second problem is addressed by using gray scale image instead of distance transform to force the watersheds coincide with the actual junctions of the image objects. The proposed procedure has been applied on the images with almost similar objects and the results of our procedure of segmentation are found to be very efficient and encouraging.