Abstract: An image fusion method based on enhancing the edges of low-frequency domain and wavelet transform is presented.First，image involved in fusion is decomposed by multi-resolution wavelet.Then the high-frequency detail images of the top layer are fused using area-based absolute maximum while the high-frequency detail images of other layer are fused using area-based maximum standard variance method.After that，an area-based feature measurement of resolution coefficient convoluted method is used for low-frequency approximation images.This method enhances low-frequency domain edge and its performance is objectively evaluated by using error of mean square root.The experiment results show that it is effective.The performance is better than the performance of the average-weighted fusion method and resolution coefficient convoluted fusion method in low-frequency domain.According to this method，the edge detail information and area-based feature of low-frequency domain are better emerged.

摘要： 提出了一种基于低频域边缘增强的小波融合方法。首先，对参加融合的两幅图像进行小波多尺度分解，然后对最高层（分辨率最低层）高频细节分量图像进行区域绝对值取大和对其它层高频细节分量图像按区域方差最大化的原则进行融合，而对低频近似分量图像采用尺度系数卷积后区域特征度量的融合方法，增强了低频域的边缘，并采用均方根误差对该方法进行了客观评价。实验结果表明该方法有很好的融合效果，与已有的低频域平均法和尺度系数卷积融合方法相比，能更好地突出低频域边缘细节信息和区域特征。