Title :
Entropy: a new definition and its applications
Author :
Pal, Nikhil R. ; Pal, Sankar K.
Author_Institution :
Electron. & Commun. Sci. Unit, Indian Stat. Inst., Calcutta, India
Abstract :
Shannon´s definition of entropy is critically examined and a new definition of classical entropy based on the exponential behavior of information gain is proposed along with its justification. The concept is extended to defining the global, local, and conditional entropy of a gray-level image. Based on these definitions four algorithms for object extraction are developed. One of these algorithms uses a Poisson distribution-based model of an ideal image. A concept of positional entropy giving any information regarding the location of an object in a scene is introduced
Keywords :
entropy; information theory; pattern recognition; picture processing; Poisson distribution-based model; conditional entropy; entropy; exponential behavior; global entropy; gray-level image; image extraction; information gain; information theory; local entropy; pattern recognition; picture processing; positional entropy; Cybernetics; Data mining; Entropy; Gain measurement; Histograms; Image processing; Image segmentation; Layout; Measurement uncertainty; Pattern recognition;
Journal_Title :
Systems, Man and Cybernetics, IEEE Transactions on
