Image encryption using a new chaos based encryption algorithm
Proceedings of the 2011 International Conference on Communication, Computing & Security
A two-dimension chaotic sequence generating method and its application for image segmentation
ICNC'06 Proceedings of the Second international conference on Advances in Natural Computation - Volume Part II
VSMM'06 Proceedings of the 12th international conference on Interactive Technologies and Sociotechnical Systems
Hi-index | 0.00 |
In this paper, a computational model for imagesegmentation based on a network of coupled chaoticmaps is proposed. Time evolutions of chaotic maps thatcorrespond to a pixel class are synchronized with oneanother, while this synchronized evolution isdesynchronized with respect to time evolution of chaoticmaps corresponding to other pixel classes in the samedata set. The model presents the following advantages incomparison to conventional pixel classificationtechniques: 1) The segmentation process is intrinsicallyparallel; 2) The number of pixel classes can be previousunknown; 3) the model offers a multi-resolution andmulti-thresholding segmentation approach; 4) Theadaptive pixel moving process makes the model robust toclassify ambiguous pixels; and 5) The model obtainsgood performance and transparent dynamics by utilizingone-dimensional chaotic maps instead of complexneurons as individual elements.