ICM And NSCT Adaptive Image Fusion Based

In recent years, Image fusion technology has been widely used because of its practical value. With the advance of various fusion methods, the fusion effect has also been improved.In order to obtain better fusion image, we studied the methods of image fusion based on multi-scale decomposition and Intersecting cortical model. Here this paper will introduce our main research findings as follows:①In this paper, we first give an overview of the meaning and methods of images fusion. Then we will highlight the principles of Contourlet transform and pulse couple neural network, and elaborate the principles and applications of Nonsubsampled Contourlet Transform and Intersecting cortical model.②Taking the characteristics of NSCT and ICM into account, here we propose an algorithm combined the NSCT and ICM, which can be applied to image fusion. Different from traditional weighted average method, the selection principle of the low-frequency subband coefficients is based on edges of images, which retained lot of edges information. The selection principle of bandpass directional subband coefficients was improved from fusion method based on ICM.③A number of parameters in ICM need to be set manually. Improved genetic algorithm is effective for optimization. It can search for the global approximate optimal solution effectively. By using the Improved genetic algorithm, the parameters can be chosen adaptively. So, an image fusion method based on ICM and NSCT is proposed, which can achieve the fusion automatically.In order to demonstrate the effectiveness of our proposed method, this algorithm has been used on several sets of different type images. And we have done many experiments to compare our image fusion methods with others (these methods are laplacian-based, wavelet-based, contourlet-based, NSCT-based, ICM and NSCT-based). The results show that our method performs a better fusion quality.