Segmentation of medical images by feature tracing in a selfdual morphological scale-space

Proceedings of SPIE - The International Society for Optical Engineering

Volumn 4322, Issue 1, 2001, Pages 139-150

  Metzler, Volker ^a   Thies, Christian ^a   Lehmann, Thomas ^a,b  

^a Medical University of Lübeck   (United Kingdom)

^b RWTH AACHEN UNIVERSITY   (Germany)

Author keywords

Causality; Feature tracing; Image segmentation; Inflection contours; Morphological scale space selfdual scale space; Multiscale approach; Reconstructive filters; Scale space analysis; Watershed

Indexed keywords

ALGORITHMS; FEATURE EXTRACTION; IMAGE ANALYSIS; IMAGE RECONSTRUCTION; MEDICAL IMAGING; OPTICAL FILTERS; STATISTICAL METHODS; TREES (MATHEMATICS);

MORPHOLOGICAL FILTERS;

IMAGE SEGMENTATION;

EID: 0034848989     PISSN: 0277786X     EISSN: None     Source Type: Journal    
DOI: 10.1117/12.431030     Document Type: Article

References (20)

1. Segmentation of textured images using a multiresolution Gaussian autoregressive model
2. A multiresolution approach for contour extraction from brain images
3. Theory of edge detection
4. Scale space filtering: A new approach to multi-scale description
5. The structure of images
6. Uniqueness of the Gaussian kernel for scale-space filtering
7. Scaling theorems for zero crossings
8. On the computation of a scale-space primal sketch
9. A multiresolution hierarchical approach to image segmentation based on intensity extrema
10. Scale-space properties of the multiscale morphological dilation-erosion
11. Gradient watersheds in morphological scale-space
12. Image segmentation and analysis via multiscale gradient watershed hierarchies
13. A multiscale gradient algorithm for image segmentation using watersheds
14. Morphological multiscale image segmentation
15. Theoretical aspects of morphological filters by reconstruction
16. Grayscale morphology
17. Stack filters
18. Morphological grayscale reconstruction in image analysis: Efficient algorithms and application
19. An efficient watershed algorithm based on connected components