Bio-inspired computing has lately demonstrated its usefulness with remarkable contributions to\r\nshape detection, optimization, and classification in pattern recognition. Similarly, multithreshold\r\nselection has become a critical step for image analysis and computer vision sparking considerable\r\nefforts to design an optimal multi-threshold estimator. This paper presents an algorithm for\r\nmulti-threshold segmentation which is based on the artificial immune systemsAIS technique,\r\nalso known as theclonal selection algorithm CSA. It follows the clonal selection principle\r\nCSP from the human immune system which basically generates a response according to the\r\nrelationship between antigens Ag, that is, patterns to be recognized and antibodies Ab, that\r\nis, possible solutions. In our approach, the 1D histogram of one image is approximated through a\r\nGaussian mixture model whose parameters are calculated through CSA. Each Gaussian function\r\nrepresents a pixel class and therefore a thresholding point. Unlike the expectation-maximization\r\nEM algorithm, the CSA-based method shows a fast convergence and a low sensitivity to\r\ninitial conditions. Remarkably, it also improves complex time-consuming computations commonly\r\nrequired by gradient-based methods. Experimental evidence demonstrates a successful automatic\r\nmulti-threshold selection based on CSA, comparing its performance to the aforementioned wellknown\r\nalgorithms.
Loading....