Image Segmentation Based on Differential Evolution Optimization

Abstract

Threshold selection is a critical preprocessing step for image analysis, pattern recognition and computer vision. On the other hand Differential Evolution (DE) is a heuristic method for solving complex optimization problems, yielding promising results. DE is easy to use, keeps a simple structure and holds acceptable convergence properties and robustness. In this chapter, an automatic image multi-threshold approach based on differential evolution optimization is presented. Hereby the segmentation process is considered to be similar to an optimization problem. First, the algorithm approximates the 1-D histogram of the image using a mixture of Gaussian functions whose parameters are calculated using the differential evolution method. Each Gaussian function approximating the histogram represents a pixel class and therefore a threshold point. The resultant approach is not only computationally efficient but also does not require prior assumptions whatsoever about the image. The method is likely to be most useful for applications considering different and perhaps initially unknown image classes. Experimental results demonstrate the algorithm’s ability to perform automatic threshold selection while preserving main features from the original image.