Download presentation

Presentation is loading. Please wait.

Published byMelissa Evelyn Modified over 2 years ago

1
Texture Featuring Based on Watershed Karsten Rodenacker, Uta Jütting GSF-Institut für Biomathematik und Biometrie Neuherberg 11.05.98

2
Content n Introduction n Some remarks about watershed n Textural featuring n Examples n Summary

3
Introduction n Textures Properties Texture elements Regularity ? What is texture ? Segmentation

4
Some remarks about watershed n Definition ä Given an image and a local maximum at point m. Consider all monotonous decreasing paths w(m,l) starting from m and ending in l. ä The set of all points l is a connected region. ä The set of all such regions is a partition of the input image, the order corresponds with the number of local maximums. ä The borders of the regions are the watersheds

5
Some remarks about watershed n Definition (continued) ä The inversion of the above definition concerning maximum by minimum and decreasing by increasing delivers a second type of watershed.

6
Some remarks about watershed n Explanation ä 1-dimensional

7
Some remarks about watershed n Explanation ä 2-dimensional

8
Some remarks about watershed n Explanation ä 2-dimensional

9
Some remarks about watershed n Result of transformation ä Input image (original, smoothed 3x3)

10
Some remarks about watershed n Result of transformation ä Watersheds (labelled regions)

11
Some remarks about watershed n Result of transformation ä Watersheds

12
Some remarks about watershed n Result of transformation ä Lower, upper ricefield

13
Some remarks about watershed n Result of transformation ä Half height segmentation HU (black)/HL (white)

14
Textural featuring n global ä Statistical estimators (moments) from: –half height segmentations and original –topological gradient ä Derived parameters –stereological (volume densities) –densitometric –morphometric

15
Textural featuring n global, e.g. volume density mean particle area Euler number Rel. density difference

16
Textural featuring n heuristical ä neighborhood related of connected regions (graph theoretical approach) –connectivity –distances ä neighborhood AND intensity related –pattern matching ä...

17
Examples n global texture features: ä VV U Volume density ä RDD Relative density difference

18
Examples n Projects ä Differentiation of hormone status of breast carcinoma patients VV U, RDD ä Differentiation of osteoblasts under different growth conditions (Osteoporosis) RHUA ä Differentiation of neuroendocrine lung tumours HUCV

20
Examples n global texture features: RDD

21
Examples

22
n Projects

23
Summary + Topology by watershed ä Taxonomy of transformation results + Segmentation + Parameter free featuring - Noise sensitivity of watershed - Difficulties of interpretation for natural images (reflected light, shadows,...)

Similar presentations

Presentation is loading. Please wait....

OK

電腦視覺 Computer and Robot Vision I

電腦視覺 Computer and Robot Vision I

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google