Abstract: Based on the redundancy and complementarity between different images of the same scenarios, a novel algorithm of image fusion based on average grads and wavelet contrast is proposed in the area of wavelet image fusion. The input images are decomposed by wavelet transform. The high frequency absolute neighborhood mean and low frequency neighborhood mean are rationed on each highest layer of each image. The bigger ratio one corresponding high frequency wavelet coefficient is taken to corresponding wavelet coefficient. While the high frequency coefficients on other decomposition level and the low frequency coefficient are calculated by gradient. The fusion image is reconstructed by wavelet inverse transform. Quantitative and qualitative analysis of the results demonstrate higher performance of the algorithm.

Key words: image fusion, wavelet transformation, average grads, wavelet contrast

摘要： 由于不同传感器获得的多幅图像对同一场景的描述具有信息的冗余性和互补性，在小波图像融合的基础上提出了一种基于平均梯度和方向对比度的图像融合方法。对参加融合的两幅图像进行小波多尺度分解，在每幅图像的最高分解层上，分别计算高频子带每个系数的邻域绝对值均值和低频系数的邻域均值之比，采用两者之比较大者所对应的高频子带系数作为融合后所对应的小波系数，对于其他分解层上的高频系数和低频系数，利用梯度最大化的融合规则得到融合图像的小波系数，通过小波重构得到融合图像。将该方法应用于仿真实验，融合图像的视觉效果有很好的改善，客观评价指标有所提高。

关键词: 图像融合, 小波变换, 平均梯度, 方向对比度