WSEAS Transactions on Systems and Control
Print ISSN: 1991-8763, E-ISSN: 2224-2856
Volume 16, 2017
A Probabilistic Model for Random Binary Image Mapping
Author:
Abstract: Many probabilistic models have been developed for numerous problems in robot and computer vision such as for image segmentation, road extraction, and object tracking. In this paper we present a probabilistic model for the random pixel mapping of binary images. The model predicts the probability of detecting dissimilarity between dissimilar binary images as a function of the number of random mappings and the amount of similarity. The model shows that detecting dissimilarity can be accomplished quickly by random pixel mapping, without the need to process the entire image. Test results on real images are presented that show the accuracy of the model.
Search Articles
Keywords: Probabilistic model, binary images, pixel mapping, image matching, image mapping, dissimilarity detection, quick matching, big images
Pages: 317-331
WSEAS Transactions on Systems and Control, ISSN / E-ISSN: 1991-8763 / 2224-2856, Volume 16, 2017, Art. #34