Parameter Adjustment of Active Models for Contour Detection with Application in AFM Images. Hwang, J., N., Torres-Tello, M., A., Rosas-Romero, R., Starostenko, O., Alarcon-Aquino, V., & Rodriguez-Asomoza, J. Journal Nanociencia et Moletronica, 6(2):1247-1262, 2008.
Parameter Adjustment of Active Models for Contour Detection with Application in AFM Images [pdf]Website  abstract   bibtex   
In the area of digital image processing, active models or snakes are mainly used for detection of object contours such as those in AFM images (Atomic Force Microscope), with certain characteristics defined by the user using prior knowledge. This information is utilized for object segmentation, object tracking, object recognition and other tasks. In this report an analysis of the main characteristics of three parametric active models (Kass, Cohen and Xu) is done. A comparative table is shown to help the user to define which could be the best model according to the application. Finally an experimental design is used to adjust the parameters of the models to guarantee a desired out put. Due to the fact that under some environments some active contour models can be recognized as being the most suitable for application, the relation among the three most referenced parametric active contour models and the selection of parameter values for each model is required for complex applications. Parameter selection is a general problem that is continuously commented in the references of this paper, and that is why a new alternative was developed to find the best parameters by means of experimentation

Downloads: 0