Fuzzy logic methods for image motion detection and analysis
Baoya Wang, C. Daly, Michael J. Ryan, M. Scott · 2002
Fuzzy logic methods have been found to be very effective in image processing. The authors (1992) previously used fuzzy logic methods to recognize the objects in two consecutive images and to compute the displacement of the corresponding object. The method used there can not compute the rotation of the object. In this paper, the concept of principal fuzzy index of area coverage (PFIOAC) is proposed and developed. PFIOAC is scale-invariant and rotation-invariant and can be used to recognize objects in two consecutive images and to compute the displacement and rotation of the object. Experimental results show that the method is effective and that it is even better to use the PFIOAC and fuzzy index of compactness approaches together.>