Processing and Image Analysis

Teorical aspects of morphological filtering have been investigated, particularly concerning the class related with morphological filters. Among other issues, the GIB has worked in combinations of filters related to the property of "strong filter" and the "levelings".

Using morphological techniques, the GIB have been investigated interpolation methods in ways that preserve and impose restrictions, if desired, the shapes of interpolated homotopy.

Image processing 01

 

Within the image processing, we have the segmentation of regions and structures of interest, which has been an active area of work which have been used, especially by means of watershed and fusion techniques of flat areas. Although the methods developed are generally applicable, have been applied especially to segment the brain structures (such as cerebral ventricles) and for extraction of particles pathology.

Following is, first, an example of segmentation and modeling of internal structures of the brain:

Image processing 02

 

And in the second place, an example of segmentation by gradient flood situations on discontinuous edges:

Image processing 03

 

References:

  1. Jose Crespo, Jean Serra, and Ronald W. Schafer. Theoretical aspects of morphological filters by reconstruction. Signal Processing, 47(2):201-225, November 1995.
  2. Jose Crespo, Ronald W. Schafer, Jean Serra, C. Gratin, and F. Meyer. The flat zone approach: A general low-level region merging segmentation method. Signal Processing, 62(1):37-60, 1997.
  3. Jose Crespo and Ronald W. Schafer. Locality and adjacency stability constraints for morphological connected operators. Journal of Mathematical Imaging and Vision, 7(1):85-102, 1997.
  4. Jose Crespo and V. Maojo. New results on the theory of morphological filters by reconstruction. Pattern Recognition, 31(4):419-429, April 1998.
  5. Javier Vidal, Jose Crespo, and Victor Maojo. A shape interpolation technique based on inclusion relationships and median sets. Image and Vision Computing, 25(10):1530-1542, October 2007.
  6. Jose Crespo and Victor Maojo. The strong property of morphological connected alternated filters. accepted for publication in the Journal of Mathematical Imaging and Vision, 2008.