IMAGE CHARACTERISTICS PROJECTS

4.3 - Color Image Analysis
The use of color may add to the value of image analysis. In order to describe and characterize color, there are a number of different color systems such as RGB, normalized RGB and IHS. The project task is to describe and compare the relative merits of at least three such color systems. Furthermore, a few color features based on the above mentioned color systems should be extracted from a set of given images from the human oral cavity. These features could then be used as input for statistical classification (see project below).

Contact: Artur Chodorowski

4.4 - Image classification
Assigning image regions or objects to classes is referred to as image classification. Given a set of image features, classification of a region can be carried out by conventional statistical methods or by non-linear methods such as neural networks. The project task is to describe some classification methods as well as to implement one of these.

Contact: Artur Chodorowski