In the field of image analysis, segmentation is one of the most importantpreprocessing steps. One way to achieve segmentation is by mean of thresholdselection, where each pixel that belongs to a determined class islabeledaccording to the selected threshold, giving as a result pixel groups that sharevisual characteristics in the image. Several methods have been proposed inorder to solve threshold selectionproblems; in this work, it is used the methodbased on the mixture of Gaussian functions to approximate the 1D histogram of agray level image and whose parameters are calculated using three natureinspired algorithms (Particle Swarm Optimization, Artificial Bee ColonyOptimization and Differential Evolution). Each Gaussian function approximatesthehistogram, representing a pixel class and therefore a threshold point.Experimental results are shown, comparing in quantitative and qualitativefashion as well as the main advantages and drawbacks of each algorithm, appliedto multi-threshold problem.
展开▼
