Genetic Neural Network In Image Segmentation

Vision is an important means of human perceiving the world. Vision information is the main source of human obtaining information. Image is the basis of human vision, it is an objective reflection of the natural world, and it is an important source of human understanding the world and humanity itself. Image processing is a technology that carried out by computer image analysis and gets the results. Image segmentation is the key technology of image processing; it is a process which can decompose an image into a lot of sets of special, non-overlapping, with a collection of strong correlation.With in-depth study of image segmentation, a variety of image segmentation methods have been poured out. In this paper, image segmentation methods are divided into the typical image segmentation method and the image segmentation method combined with some specific theories. The former contains the threshold method, the regional method, the edge detection method and the clustering method, the latter often combines the typical image segmentation methods with fuzzy sets, genetic algorithm, neural network, and rough set. This paper proposes a method of image segmentation that based on improved genetic neural network. The paper also lists out the SM, UMA, UM, GC and time complexity five kinds of evaluation criteria, evaluates the methods of image segmentation from both subjective and objective. BP neural network and genetic algorithm is the theoretical basis of the paper. BP neural network is a multilayer feed-forward neural networks; it has wide adaptability and effectiveness, and it is a manifestation of the essence for artificial neural networks. Genetic algorithm is an efficient global and heuristic optimization method which is proposed by simulating natural evolution of biological, it has advantages such as simple algorithm, parallel operations, global searches, etc.In this paper, image segmentation is regarded as a classification problem, so how to get an efficient classification method is the focus of this study. Genetic neural network is a common method to solve classification problems, it is a network model that use genetic algorithm to optimize the BP neural network. In order to enhance network performance, the fitness function and genetic operators of genetic neural network are improved in this paper, and the simulation results show that the training speed and convergence precision of improved genetic neural network are greatly enhanced compares with traditional genetic neural network.In this paper, the objects of segmentation are the gray images with a specific goal, and the paper sets the target sample by image histogram analysis, learns the sample with improved genetic neural network, gets the trained network, then makes the pixel matrix of image as the input vectors, put the input vectors into the trained network to classify, achieves segmentation at last. The simulation results prove that the method of image segmentation based on genetic algorithm approach a good segmentation compares with the traditional methods of image segmentation, it is a feasible method of image segmentation.