Title :
Cross-entropy Histogram Equalization
Author :
Roopaei, Mehdi ; Agaian, Sos ; Shadaram, Mehdi ; Hurtado, Frank
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of Texas at San Antonio, San Antonio, TX, USA
Abstract :
This paper introduces a novel cross-entropy Histogram Equalization method. The new algorithm is based on separating the density spectrum of an image and then minimizing the so-called Cross-Entropy between the brightness and darkness components. Cross-entropy is used to measure the information-theoretic distance between the image´s brightness and darkness. Experimental results show that the cross-entropy histogram equalization method has better performance than Bi-histogram equalization and its modified weighted form.
Keywords :
entropy; image enhancement; cross-entropy histogram equalization method; image brightness; image darkness; image density spectrum separation; information-theoretic distance; Boats; Brightness; Consumer electronics; Entropy; Histograms; Image enhancement; Transforms; Brightness/Darkness decomposition; Cross-entropy; Histogram Equalization; Image enhancement;
Conference_Titel :
Systems, Man and Cybernetics (SMC), 2014 IEEE International Conference on
Conference_Location :
San Diego, CA
DOI :
10.1109/SMC.2014.6973900
