Title :
A Kind of Two-Dimensional Entropic Image Segmentation Method Based on Artificial Immune Algorithm
Author :
Li, Youxin ; Mao, Zongyuan ; Tian, Lianfang ; Tan, Guangxing
Author_Institution :
Coll. of Autom. Sci. & Eng., South China Univ. of Technol., Guangdong
Abstract :
Two-dimensional entropic segmentation method has been greatly developed because of high segmentation accuracy and good stability, while a hard problem is that it gives rise to the exponential increment of computational time in comparison with the traditional one-dimensional histogram partition technology. To solve this problem, a new kind of image thresholding method is presented based on the combination of the artificial immune algorithm (AIA) and two-dimensional entropy techniques in this paper. This method can effectively improve the computation time and avoid getting into local optimization of the threshold by making use of AIA´s characteristics of the intelligent computation, adaptive evolution and globally optimizing. The test results show that the method is effective and practicable
Keywords :
computational complexity; entropy; image segmentation; optimisation; 2D entropic image segmentation; adaptive evolution; artificial immune algorithm; global optimization; image thresholding; intelligent computation; Automation; Educational institutions; Entropy; Histograms; Image segmentation; Optimization methods; Partitioning algorithms; Pixel; Stability; Testing; Artificial immune algorithm; Image segmentation; Two-dimensional entropy;
Conference_Titel :
Intelligent Control and Automation, 2006. WCICA 2006. The Sixth World Congress on
Conference_Location :
Dalian
Print_ISBN :
1-4244-0332-4
DOI :
10.1109/WCICA.2006.1714043