Defuzzification 2015 - Special Session on Defuzzification: When discrete geometry meets fuzzy geometry

Image data is inherently discrete since it consists of a finite number of pixels, where each of them has a single intensity. In this case, standard digital concepts, such as distance and lines, need to be defined discretely. Many researches are directed towards discrete geometry in image processing (segmentation, objet recognition, etc.).
Alongside, in fuzzy set theory, each pixel has a degree of membership to different sets, as opposed to crisp sets, where each pixel belongs to only one. This principle is used in particular in the image analysis process for images with noise and artefacts. Fuzzy logic uses aspects of geometry for modeling membership functions of fuzzy sets. For instance, most of them take the form of piecewise linear functions. Defuzzification of a fuzzy image can be a good alternative to conventional segmentation methods, and has been proven to give good crisp segmentation results in various applications.
The aim of this session is to provide a presentation of recent results and advances of how discrete geometry can aid in the decision (defuzzification) process when working with image data. The session will be an opportunity for researchers in this area to come together, and a way to establish international collaborations in such interesting field, concerning theoretical and applicative researches. Topics of this special session are thus highly related to the IPTA conference, and include, without being exhaustive:
Fuzzy sets
Type-I fuzzy sets and type-II fuzzy sets
Defuzzification
Discrete/digital geometry
Mathematical morphology
Image enhancement
Segmentation and edge detection
Feature extraction and pattern recognition
Medical image processing
Convexity and applications (i.e. Helly’s theorem)
Polynomial methods and applications