自适应PCNN的形态小波多聚焦图像融合方法  

Fusion algorithm of multi-focus images based on morphological wavelet of adaptive PCNN

作　　者：何刘杰[1,2] 胡涛[2] 任仙怡[2] 

机构地区：[1]深圳大学计算机科学与技术系,广东深圳518060 [2]深圳信息职业技术学院,广东深圳518029

出　　处：《计算机工程与应用》2013年第12期132-135,159,共5页Computer Engineering and Applications

基　　金：国家自然科学基金(No.60971120);广东省自然科学基金(No.9151009001000052)

摘　　要：为了解决传统形态小波图像融合方法在重构尺度信号时发生了位置错误和重构细节信号时发生了灰度值下溢的不足,提出一种有效的基于自适应脉冲耦合神经网络(PCNN)的形态小波多聚焦图像融合方法。通过形态小波对已配准的源图像进行分解;提出一种自适应的PCNN,用分解系数的改进拉普拉斯能量和(SML)作为PCNN对应神经元的反馈输入,用图像的清晰度作为对应神经元的连接强度,经过PCNN点火获得参与融合系数的点火映射图,通过判决选择算子指导系数的融合;经过形态小波逆变换得到融合图像。实验结果表明,该算法的融合图像具有良好的视觉效果及较高客观评价指标。A fusion algorithm of multi-focus image fusion is proposed based on morphological wavelet of adaptive Pulse Coupled Neural Networks（PCNN）. Two original images are decomposed by using morphological wavelet separately, thus the low frequency subband coefficients and varieties of directional bandpass subband coefficients are obtained. This algorithm uses the sum-modi- fied-Laplacian of each pixel as the value of the feeding input of each neuron and uses the contrast of each pixel as the value of the linking strength of each neuron. After the processing of PCNN, new fire mapping images are obtained for each coefficient. The clear objects of each original image are decided by the compare-selection operator with the fire mapping images pixel by pixel and then all of them are merged into a new clear image. The fused image is obtained by performing the inverse morphological wavelet on the combined coefficients. The experimental results show that the proposed algorithm outperforms traditional fusion algorithms in terms of objective criteria and visual anoearance.

关 键 词：多聚焦图像融合 形态小波 脉冲耦合神经网络 改进拉普拉斯能量和 清晰度 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TN911.73[自动化与计算机技术—计算机科学与技术]